I was dying. Not literally, of course, but my body felt like it was giving up. My head felt heavy, my eyes were streaming and my sinuses held what seemed to an industrial supply of snot. I had the flu.

In a ‘normal’ job, I could take a sick day or three and properly rest and recover, because I’d still be getting paid. But as a content creator, I didn’t have that luxury. I had a video to get out, short-form content to publish, and self-marketing to complete. I simply couldn’t afford to stop, because doing so would spell death to my 100,000-subscriber YouTube channel.

So I had to load myself up on medicine and crack on. But these days, AI tools mean that you don’t have to do the same.

Quick summary: which AI voiceover platform wins?

If you’re looking to create a faceless channel or a digital clone of yourself, HeyGen’s suite of tools will suit you very well, especially given that it can translate your content into 175 different languages. For most creators, though, ElevenLabs’ advanced AI voiceovers will be the better choice. The platform can fix your stumbles, as well as expand out your product portfolio to audiobooks and podcasts. It’s an essential part of the modern creator’s toolkit.

ElevenLabs is an advanced AI platform specialising in creating a clone of your voice. For me, struggling to stay upright at my desk, this would have been a godsend, because I could have used my AI voice to finish my next video. It has quickly become the gold standard among creator teams looking to make sure that when their lead presenter can’t speak — or isn’t available — that their channels don’t suffer as a result.

HeyGen goes one step further, creating a digital clone of yourself so that you can be the director of your videos, rather than the presenter. And for anyone looking to start a ‘faceless’ YouTube channel in 2026 — a market which can generate six-figure incomes for the most successful creators — could be your fast-lane ticket to success.

Both offer something different, with ElevenLabs focussing on AI voices and HeyGen offering full avatars. But here’s a secret most people won’t tell you — HeyGen actually relies on ElevenLabs itself to produce the voices of its AI avatars. So by choosing HeyGen you could ultimately get the best of both worlds.

In this article, I want to explore which one would have been the perfect fit for my ‘illness’ workflow back in the day, and which one offers the best blend of features for today’s creators. I’ll go through the features, cost and ease of use of both platforms before offering my recommendation. For this article, where a key part of my conclusion will come from the effectiveness of both platforms in terms of their final products looking and sounding right, I’ll also add an Uncanny valley score.

Ready to start? Let’s dive in.

Before we start: How I rate and review products and services

As a former creator with an audience on YouTube and TikTok which at one point totalled some 200,000 people, I don’t just look at specs — I look at the sanity-saving measures offered by each product. And to help you decide whether ElevenLabs or HeyGen is best for you, I’ll be breaking my analysis down into three key areas: features, cost and ease-of-use.

Features: I’m looking for the tools to give creators that valuable edge in what I know is an increasingly crowded market. These tools should solve real problems, and I’ll prioritise those tools which allow AI to do the heavy lifting.

Cost: I’ll break down what each program offers in its free tier, as well as the extra services you get as a paid member. We’ll look at both the monthly and annual cost, and I’ll give you my verdict on whether either product is worth paying for. I won’t recommend paying if I don’t believe in the product, because as a former creator, I know that money is often tight for non-monetised channels.

Ease-of-use: I’m not just looking for tools, I’m looking for the workflows which could save a burned-out creator. A winning program here will offer a simple user interface with easy-to-find tools, and a learning curve which isn’t overwhelming.

Uncanny Valley: How much does the final AI product look or sound like me?

ElevenLabs — the AI voice architect

Features

If you empathise with the feeling of having to record a video while also suffering from illness, then you’ll appreciate ElevenLabs’ flagship feature — a digital, AI-driven clone of your voice. Think of this as your digital twin, able to step in when you’re unable to. And writing this, I just wish that ElevenLabs was around when I was on my 100k grind.

What sets ElevenLabs’ voice cloning abilities apart from the competition — and there’s a growing amount of that — is the sheer depth of its voice training. While some platforms will claim to be able to generate a passable voice clone based on 60-seconds of training, ElevenLabs requires substantially more data.

The process of training an ElevenLabs voice takes about 30 minutes of high-quality audio, which being a creator, you likely already have on hand from previous videos. If that sounds like a bit of a faff, remember that the resulting voice truly sounds like you. And once trained, you can use it forever.

I’ve tried out this feature myself, based on voice training data from my old videos. And I was astounded with the result. The voice not only sounds like me, but comes with the same speech patterns, sentence rhythm, pitch and nuance with which I speak. It sounds eerily human, and if I didn’t know better, I’d assume it was.

If you prefer speaking to writing, as plenty of creators will, then you don’t need to write down your scripts — instead, you can use ElevenLabs’ speech-to-speech function and simply say what you want your digital twin to say. So in my case, my croaky, hoarse, flu-ridden voice could speak what I wanted my non-ill clone to say for me. And the result is a healthy voice which actually sounds like me.

Now ElevenLabs has come a long way over the past few years, and its latest V3 engine — launched in May 2026 — understands context and subtext better than ever. That means you can give your AI voice clone direction, telling it how to perform the lines you’ve given. Direction tags such as ‘whispers’, ‘sighs’ or ‘shouts’ will result in a dramatically different performance. So you can create the kind of emotion you’d like to see in a Hollywood film, on a straight-to-DVD budget. The system is even smart enough to make the end of a sentence sadder without being told to do so, if the context demands it.

And since just one-in-five of YouTube’s audience speaks English, ElevenLabs’ ability to automatically re-dub your voice clone into 29+ different languages — including French, German and Japanese — could help to bring your content to wider audiences and unlock new revenues for your channel. And crucially, even in different languages, ElevenLabs’ can keep the original emotion of your voice. I can’t over-estimate how powerful that is.

Other features include ‘flows’ which act like AI agents and can automatically complete tasks once you set them going. This could be useful if you plan on using ElevenLabs to create lots of the same kind of content. Plus, while not its core focus, ElevenLabs can also create image and video content using text prompts, which appear to be powered by ChatGPT.

FEATURES SCORE: ⭐⭐⭐⭐⭐(5/5)

Cost

As I’ve said in previous articles, any platform which offers a free tier which could be helpful for smaller creators gets extra points in my book, and ElevenLabs falls straight into that category.

Its free tier gives you 10,000 characters to play with, which is enough for about 10 minutes of audio per month. And that means you could use your digital twin to cover mistakes or generate the odd sentence without ever paying a penny for the platform.

If you want access to the most advanced features, though, including instant voice cloning, you’ll need to upgrade to at least ElevenLabs’ Starter tier. This costs $6 per month, which is very reasonable in my book, and crucially adds commercial rights to your generated audio.

The Creator plan costs from $22 per month and is my recommended choice, especially for larger channels. This unlocks the professional-grade voice cloning I mentioned in the features section above. It also gives you enough credits to generate about 1.6 hours’ worth of audio per month, which based on an average length of 10 minutes, is enough for about nine videos. It allows you to start a faceless channel and keep it going, which in 2026 is a killer move.

ElevenLabs’ most expensive tier aimed at individuals is the Pro plan, which at $99 per month is a hefty investment. But for those planning on producing huge amounts of audio content — think daily podcasts or audiobooks — it could be worth the investment. You get 500,000 characters of recording and priority rendering.

There are also additional tiers, dubbed Scale and Business, which are geared towards creator teams, and which cost $299 per month and $990 per month respectively. Larger teams can opt for an Enterprise subscription with a custom price.

COST SCORE: ⭐⭐⭐⭐⭐ (5/5)

Ease-of-use

ElevenLabs’ user interace is dark and easy to navigate — things I love, given I’d usually be looking for productivity tools in the early hours of the morning.

The layout makes it easy to see what you’ve got access to and what you might want to do, whether that’s create a dubbed video or voice an audio book.

I like that there are clear options to choose voices other than your own — which might be useful if you plan on running multiple faceless channels — as well as training your own voice.

I also like that the most advanced features aren’t buried away in menu after menu. For example, clicking on ‘create voice’ opens up another box for you to then choose how that voice will be made, either through text prompts, through instant cloning or professional cloning (on the Creator plan). Or you can use text prompts to remix ElevenLabs’ extensive library of voices.

And if you make a mistake, the AI can fix it for you. For example, you don’t have to re-record an entire line if one word is out of kilter, you can fix it with a text prompt.

The learning curve is far shallower than with some AI tools I’ve reviewed in the past, and is as much of a plug-and-play system as it’s possible to get.

Quick tip: Elevenlabs sliders

When creating your voice, you’ll see two sliders for stability and similarity. These govern how strictly the AI voice must stick to your instructions. Most people crank these all the way up to 100% and never touch them again, but I’d advocate giving the AI voice some breathing room, which allows for a wider variety of emotion. My preset kept stability to 40% and similarity to 75%.

EASE-OF-USE SCORE: ⭐⭐⭐⭐⭐(5/5)

HeyGen — your digital twin

Features

Sometimes, simply having a workable voice isn’t enough to get your video over the line.

HeyGen’s headline feature looks to remedy that situation, by creating a fully digital twin which can act as your stand-in when you’re under the weather, or when your social battery is just drained. It’s exactly the kind of feature I would have wished for sat dead-eyed at my desk, dreading the thought of pressing record.

It works by analysing uploaded footage and creating your avatar from it. And the chances are that as content creators, most users are likely to have an extensive library of footage featuring themselves to choose from.

Quick tip: How to create the best digital twin with HeyGen

You can’t just upload any old footage to HeyGen and expect a fully formed digital twin out the other side. The footage you upload needs to be at least two minutes long, feature only one person in front of the camera, and in a well-lit environment.

The process works best when you show a range of expressions in your uploaded footage, including plenty of mouth movement. Keep yourself in the centre of the frame so that you don’t move around too much, and check that the camera is stable. This ensures the best results.

Once you have your digital twin in place, you can put them to work in a variety of formats. HeyGen has templates for you to choose from, or you can create your own content in both portrait and landscape formats.

You select your digital twin from the list of availble avatars, and write what you want them to say — or upload and record audio which you want your avatar to re-voice. And as I mentioned before, HeyGen effectively plugs into ElevenLabs, using the latter’s powerful voice cloning tool to give its avatars the best possible sound.

The result is incredibly powerful, with the finished avatar not only able to convincingly move, look and sound like you, but it’s able to be placed in a variety of different backgrounds and lighting conditions. In one example I saw, a created avatar repeated the same script first in front of a brightly lit building, then in a dark gaming room, and then in a more neutral setting.

HeyGen has been hooked up with Sora 2 and Veo, allowing you to create custom digital backgrounds for your avatar.

Of course, there’s some debate to be had over whether regular viewers would be able to tell the difference between the ‘real’ you and your digital twin, and under what circumstances you might wish to disclose the difference. But thinking back to my own use case, I would have loved my digital twin to take over when I was under the weather.

One of HeyGen’s other prime features is the ability to lip-sync your mouth movements to match whatever language your video is presented in. To put that in real terms, consider this: I cannot speak French, German or Spanish, but my digital twin can. And he’s damned good at it, allowing me to put my content out to other countries and potentially earn from views received there.

FEATURES SCORE: ⭐⭐⭐⭐⭐(5/5)

Cost

Like ElevenLabs, HeyGen offers a free tier which could prove to be all that most single-person YouTube channels would need. The free tier allows you to create three videos per month, with each video lasting up to one minute. You can export in 720p and can choose from a catalogue of more than 500 pre-made digital twins. Plus, you get access to dubbing in 30+ languages.

The Creator tier costs $28 per month or $295 per year. Crucially, this allows you to export your finished videos in 1080p, plus you can record up to 30 minutes at a time with 200 monthly credits. You get faster video processing, the ability to remove the HeyGen watermark from your videos and download them, plus unlimited photo avatars.

The language roster expands to 175+ languages, and you get additional features including voice cloning, the ability to generate text and video within HeyGen, and your unused credits roll over each month.

The highest tier is the Pro plan, which costs $100 per month or $965 per year. That sounds like a big investment, obviously, but you also get access to the most advanced features. Those include the ability to export your finished videos in 4k, faster processing, and the ability to create your custom ‘digital twin’ in glorious 4k resolution.

I dare say that most creators will find everything they need in the Creator plan, while smaller channels or those who want to try out the HeyGen platform without committing to a recurring financial cost will be impressed with what you get for free.

If you do decide to pay more, think of it as an investment in a digital actor which could be making you money for years to come. And if you think about it in terms of replacing the cost which you might otherwise need to put into a camera, sound and lighting, the value quickly adds up.

COST SCORE: ⭐⭐⭐ (3/5)

Ease-of-use

For creators who imagine themselves to be the next Christopher Nolan or Steven Spielberg, HeyGen’s interface will be right up their street, because you act as a director for your digital twin.

The interface to create your videos with HeyGen takes a little getting used to, because you work primarily with a script-driven timeline. However, the redesign given to the platform in 2026 has introduced a more free-from approach (similar to Canva) where you can visually drag, drop, and arrange your avatar, B-roll, and text overlays.

What truly saves your sanity is the new video agent feature. Instead of manually cutting every scene, you can describe your vision to the AI assistant — for example“Make a 60-second TikTok about GTA RP trends with a cinematic background” — and the agent handles the first draft, selecting the best camera angles and layouts for you.

When I was making content, I was the presenter. With HeyGen, I am instead the director. You no longer have to worry about lighting your room or whether your hair looks right; you’re managing a dgital asset that is always camera-ready. It’s a refreshing change for anyone on the brink of burnout.

EASE-OF-USE SCORE: ⭐⭐⭐⭐(4/5)

ElevenLabs vs. HeyGen: the verdict

ElevenLabs and HeyGen take a different approach to helping creators who want to have some — or all — of their presence in front of the camera augmented by AI. And thinking back to those times when I didn’t have the energy to finish the video I was working on, both would have been very useful for me.

But before I deliver my final verdict on which one best deserves your time and money, let’s look at the pros and cons of both services.

ElevenLabs: pros and cons

PROS

• Incredible realism: ElevenLabs’ voices are industry leading in their clarity.
• Speech-to-speech direction means you don’t even have to write down what you want your voiceover to say.
• Generous free tier and high-value starter plan is cost-effective for money-conscious creators.

CONS

• Audio-only synthesis means you’ll need another tool (like HeyGen) to create a full avatar.
• Professional voice cloning requires 30 minutes of high-quality audio to use as a sample, and training can take several hours.
• If you’re producing lots of content, your quota of credits could run out quickly.

FINAL SCORES

• Features 5/5 ⭐⭐⭐⭐⭐
• Cost 5/5 ⭐⭐⭐⭐⭐
• Ease of use 5/5 ⭐⭐⭐⭐⭐
• Uncanny valley 5/5 ⭐⭐⭐⭐⭐

Best for: ElevenLabs is best for tired (or ill) creators who still want their videos to go out on time, with minimal interruptions.

HeyGen: pros and cons

PROS

• Creates a fully digital stand-in for your content, opening up the possibility of a faceless channel.
• Ability to lip-sync your content into 175+ languages, potentially boosting revenues.
• ‘Director’-style workflow allows you to customise your scenes with high precision.
• Integration with industry-leading platforms including ElevenLabs, Sora and Veo.

CONS

• Steep potential investment on paid tiers.
• A steeper learning curve than with ElevenLabs.
• Rendering 4K video can take a long time, especially on lower tiers.

FINAL SCORES

• Features 5/5 ⭐⭐⭐⭐⭐
• Cost 3/5 ⭐⭐⭐
• Ease of use 4/5 ⭐⭐⭐⭐
• Uncanny valley 4/5 ⭐⭐⭐⭐

Best for: HeyGen is best for creators who want to replace themselves entirely with a fully AI-driven digital twin who can shoulder the heavy burden of presenting.

CONCLUSION

Let me clearly address the question asked in the headline to this article. Are content creators obsolete? The answer is no, the creators themselves aren’t, but some of their gruntwork is.

If you feel like rubbish and can’t face the camera, both of these tools can provide you with a way out, while still making content. I ultimately think that more creators will be best served by ElevenLabs, since it allows you to continue your workflow while still placing the thing that made your channel great in the first place — you — front and centre.

If, on the other hand, you suffer from the understandable anxiety which comes from putting yourself in front of the camera, or simply want to switch to a faceless channel experience, then HeyGen is well worth considering. For creators who never want to step in front of the camera themsevles, it provides a route to still making human-centric content.

Want to read more? My deep dive into Descript (the ‘brain’ of my setup) is already live here. Check it out to see how I handle the long-form side of things. Or if you want further comparisons, why not check out my verdict on OpusClip vs Munch Studio.

Note: This article contains affiliate links. If you click and buy, I may earn a small commission at no extra cost to you. I only recommend tools I wish I’d had during my 100k journey.

Get your tags right. You can have a cracking video, but if your YouTube description and tags aren’t catchy, you’re going to fail.

That advice was the strongest I received when I was grinding on my virtual reality gaming channel. I’d just hit 10,000 subscribers and was starting the long climb to 100,000. And I just couldn’t get it right. I saw my rivals creating what I saw to be videos with less storytelling than my own, but they consistently performed better because they understood the value of tagging, and I didn’t.

Quick summary: which AI co-pilot wins?

If you want the best data available which can give you actionable insights on your channel from day one, choose VidIQ. If, however, you want access to a large library of stock footage and are looking to build automation into your channel in a bid to save time, choose TubeBuddy. For my money, I’d value knowing with greater certainty that my planned video would perform well, so I’d choose VidIQ.

I spent hours on the 2024 versions of TubeBuddy and VidIQ, frantically researching the best tags to use in the hope that they’d pull up my stalling channel. But my decision fatigue had already set in, and by the time the fog had lifted, my drive had gone. I did eventually reach the 100,000-subscriber mark, but not long after I shut up shop for good. I was simply too tired to continue.

In 2026, the same advice I clung onto so dearly doesn’t mean anything any more. YouTube’s algorithm has moved beyond simple tags and descriptions to something altogether more advanced. Its AI systems can detect what your video is about and categorise it far more accurately than with simple tags. And data-driven ideation is now king. Knowing your ‘outliers’ — the videos in similar niches where your competitors have flourished — is paramount to success.

It follows, then, that both TubeBuddy and VidIQ have also evolved to offer far more than simple SEO research.

Both are now paradigms of productivity, but for very different reasons. In this article, I’ll explore what each offers from a creator’s perspective, and give you my recommendation as to which is ultimately worth your time — and money.

Before we start: How I rate and review products and services

As a former creator with an audience on YouTube and TikTok which at one point totalled some 200,000 people, I don’t just look at specs — I look at the sanity-saving measures offered by each product. And to help you decide whether TubeBuddy or VidIQ is best for you, I’ll be breaking my analysis down into three key areas: features, cost and ease-of-use.

Features: I’m looking for the tools to give creators that valuable edge in what I know is an increasingly crowded market. These tools should solve real problems, and I’ll prioritise those tools which allow AI to do the heavy lifting.

Cost: I’ll break down what each program offers in its free tier, as well as the extra services you get as a paid member. We’ll look at both the monthly and annual cost, and I’ll give you my verdict on whether either product is worth paying for. I won’t recommend paying if I don’t believe in the product, because as a former creator, I know that money is often tight for non-monetised channels.

Ease-of-use: I’m not just looking for tools, I’m looking for the workflows which could save a burned-out creator. A winning program here will offer a simple user interface with easy-to-find tools, and a learning curve which isn’t overwhelming.

VidIQ: What are outliers and why are they so important?

Features

I explained in the introduction to this article that in 2026, YouTube’s algorithm has moved beyond titles and tags. Now, it’s looking for videos which have the capability to dramatically out-perform other content on your channel. Those videos are known as outliers — and VidIQ’s killer feature is being able to spot them.

This is the feature I wish I’d had during those long nights trying to picture what my next video should be. And the capacity to get this wrong is high, as I well know.

To give you a real-world example, one of my best-performing videos was a VR comedy video about the viral video game, Among Us. It out-performed anything else I’d ever done. So, naturally, I followed it up with a video about a non-VR game in a completely different genre. Mad, right? Well, if I’d have access to VidIQ at the time, I wouldn’t have made that mistake.

So, how does it work? VidIQ scans content within your niche to find the videos which have massively out-performed a channel’s usual content and average subscriber count. If a channel with 5000 subscribers suddenly produces a video which racks up 500,000 views, then that’s an outlier. VidIQ will store those videos so you can see if you can repeat their success. I don’t mean copy them, of course, but spark your own creativity with a new idea.

I especially liked chatting with VidIQ’s embedded AI bot, which acts as a sounding board for your ideas. Even for one-person channels, it gives you the feeling of having a full team around you. That AI bot can help with all aspects of ideation, from finding the initial idea, to helping with scripting, hooks and even thumbnail concepts. And it’s all backed up by real-time data on what’s trending at the moment on YouTube.

Using its historical data, VidIQ gives your next video ideas a virality score, which acts as an insurance policy against spending 20-plus hours on a video nobody is going to watch. Like I did all those months ago…

Even better, you get to start each morning with tailored list of up to 50 video ideas. And the chances are that at least some of them will be bangers. Even before pressing record, then, you’re giving your video the best chance at hitting the algorithm’s sweet spot.

FEATURES SCORE: ⭐⭐⭐⭐⭐(5/5)

Cost

VidIQ comes in clutch with a free tier which, as an entry-level creator or one who is just getting started, gives you a great taste of what the platform can do. You get basic keyword research and three daily video ideas, but you’ll quickly find yourself hitting that limit.

The paid tier, called Boost, starts at a reasonable £19 per month, or £199 per year, and gives you more keywords and the ability to track them over time. However, to access the advanced AI features we looked at in the previous section, you’ll need to upgrade to the Max plan.

At £39 per month or £372 per year, VidIQ’s Max plan is a hefty cost, but for channels who either want a headstart on the competition or who plan on uploading regularly, it’s a small price to pay. For a channel sitting at 10,000 subscribers which wants to make the jump to 100,000 and that vital goal of monetisation, I’d say it should substantially shorten the journey.

For me, it’s the sanity tax that’s worth paying to get as close to guaranteeing consistent growth as it’s possible to get.

COST SCORE: ⭐⭐⭐ (3/5)

Ease-of-use

I’ll always award extra points for a program whose default setting is dark mode — a must for the perpetually tired creator. And as well as being easy on your eyes, VidIQ’s user interface looks sleek and packed with features. Staring at it gives you the same feeling as a Wall Street trader looking at their data-loaded terminal.

I also like the VidIQ doesn’t require you to log into its platform in a new window, because it lives as a browser extension and as part of your existing YouTube Studio. That functionality is great, because it puts VidIQ’s features in the context of your existing channel. And since they’re right in front of you when you’ll be using YouTube Studio in any case, you won’t forget to take full advantage of them.

VidIQ’s browser extension allows you yo see your real-time stats as you work, which is incredible for your workflow. That said, because the program gives you so much data, it can feel a bit overwhelming. That said, if you’ve paid for the Max plan, then VidIQ’s AI co-pilot does a good job of surfacing the most useful stats for you.

A quick note on performance, too, because this is one hefty browser extension, so if you’re gaming, editing and using VidIQ on a fairly old machine, you might see it start to struggle.

EASE-OF-USE SCORE: ⭐⭐⭐⭐(4/5)

TubeBuddy: the total tool-chest for tired creators

Features

To put these two programs into context, if ViqIQ is your data scientist, giving you actionable advice based on reams of numbers, then Tubebuddy is your production manager, able to take the often crushing weight of running your channel off your shoulders.

It’s a burden I know only too well. Once I’d ideated my next video, I had to record and edit it, create the social assets, design the thumbnail, do my SEO research and promote the new video on my social channels once I’d hit publish. It was exhausting and contributed to the burnout which eventually stalled my channel growth. Had I trusted TubeBuddy at the time, I might still be making content today.

Among TubeBuddy’s tool set is A/B thumbnail testing. You’ll see larger creators test their thumbnails using exactly the same principle — throw out two versions, and choose the one which resonates with your audience the best. TubeBuddy removes that guesswork by automatically testing your thumbnail concept, allowing you to make a more informed decision on which one will get you the biggest click-through rate (CTR). If you’re spending money on a thumbnail artist, this means you only need to commission one design rather than two.

Then there’s the ability to bulk process the admin work of your channel. I remember the joy of securing my first sponsor, who wanted a small sentence including at the bottom of each video description. I had to update the description on some 100 videos individually, and it took me a whole afternoon. TubeBuddy would have allowed me to do the same job in minutes.

And if you find yourself with gaps to fill in your video, TubeBuddy’s library of stock footage and audio is one of the best you’ll find anywhere. I was forever needing to find obscure footage to support my VR comedy skit videos, and it was a seriously long process which TubeBuddy could have helped avoid. I also like tha TubeBuddy includes a ‘Creator News’ feed which puts the biggest headlines and changes from the creator world directly into your dashboard.

Finally, while VidIQ scores outliers, TubeBuddy gives you a more long-term look at your channel health, including how you’re stacking up against key rivals in terms of views and subsriber count over 30, 60 and 90 days.

FEATURES SCORE: ⭐⭐⭐⭐(4/5)

Cost

Like VidIQ, TubeBuddy’s free tier is very much a taster of what the paid tiers offer, but for new creators it’s a great place to start. You get basic tag suggestions and some productivity tools, but the real bulk-editing time-savers are locked away unless you pay for them.

Fortunately, you don’t need to pay much to unlock those tools. TubeBuddy’s Pro plan costs £11 per month – which is one of the best-value packages in the industry. And it gives you most of the features you’ll want, including unlimited keyword research and a solid suite of SEO tools. If your channel is sitting at 1000 subscribers and you’re wondering how to take the next step, it’s well worth considering.

The aptly-named Legend plan starts from £24 per month, and is what you’ll need if you want to unlock the advanced A/B thumbnail testing I described earlier. This is geared towards channels targetting serious growth.

For creator teams with multiple channels, TubeBuddy offers an enterprise tier with 1:1 training sessions and priority service, but you’ll need to get in touch with them for a price.

If it was me, I’d go straight for the Pro plan. For not much outlay you get some incredible tools. And I’m not getting paid to say that, either, since TubeBuddy isn’t even accepting new affiliates at the time of writing.

COST SCORE: ⭐⭐⭐⭐⭐(5/5)

Ease-of-use

Like VidIQ, TubeBuddy runs within the existing YouTube Studio, meaning it’s well integrated (but also that you might run into the same performance issues if you’re using ageing hardware.) And again, huge plus points for having an integrated dark mode as the default — my tired eyes are grateful.

I prefer VidIQ’s relatively sparse, modern layout over the relative clutter of TubeBuddy. Some may like having every data point within their finger’s reach, but I found it somewhat overwhelming, with the interface feeling rather busy by comparison. Trying to find what you want can lead you down a rabbit warren of menus. It took me a few days to learn the ropes, and the learning curve is steeper than that of VidIQ.

It’s worth noting that both VidIQ and TubeBuddy offer mobile apps, but the latter’s feels rather limited in 2026. You can check your channel stats on the go, but not much more than that. You’ll want to return to your desk to access the best features.

EASE-OF-USE SCORE: ⭐⭐⭐⭐(4/5)

VidIQ vs TubeBuddy: the verdict

TubeBuddy and VidIQ both offer tools which could have staved off the burnout which ultimately ended my own creator journey. And to be honest, I would have made good use of both of them at different stages in my channel’s growth.

But before I deliver my final verdict, let’s look at the pros and cons of both services.

VidIQ: pros and cons

PROS

• Outlier detection is an incredibly powerful tool for video ideation
• AI co-pilot gives one-person teams a crucial sanity check
• Sleek, modern interface that’s easy to understand

CONS

• Best features are locked into the most expensive tier
• Extension could run poorly on older machines
• Works best for finding new ideas, rather than managing your existing roster of content

Best for: creators who are comfortable with where their channel is at today, but want to secure growth for tomorrow.

TubeBuddy: pros and cons

PROS

• A/B thumbnail testing can save time and money
• Bulk processing of your existing videos
• Huge stock library of both video footage and audio

CONS

• More cluttered user interface
• Some tools feel a bit dated in 2026
• Mobile app isn’t useful for much more than checking your stats

Best for: creators with a large library of existing content, and no way to manage it effectively.

CONCLUSION

As I’ve mentioned above, the answer to the question of ‘which of these AI-powered co-pilots could have saved my channel’ is unusually nuanced here, because it’s both of them. TubeBuddy would have helped me to take control of my growing library of content, while VidIQ would have kept me focussed on the next banger video.

If the biggest problem facing your content right now is decision fatigue, then VidIQ will be magical to you, because it can reinforce the direction your channel is taking.

Meanwhile, if your biggest barrier to growth is productivity, TubeBuddy’s suite of tools can help to bring your content under control and quickly roll out changes — useful if you suddenly find yourself landinfg your first sponsor.

If you can only choose one, my recommendation is VidIQ. For YouTube’s current algorithm, where a viral idea is your biggest asset, it makes sense to get to know what makes an outlier video, and turn your next upload into one. And at the end of the day, I’d rather have the data to pick a winner than the tools to manage a flop.

Want to read more? My deep dive into Descript (the ‘brain’ of my setup) is already live here. Check it out to see how I handle the long-form side of things. Or if you want further comparisons, why not check out my verdict on OpusClip vs Munch Studio.

Note: This article contains affiliate links. If you click and buy, I may earn a small commission at no extra cost to you. I only recommend tools I wish I’d had during my 100k journey.

Every creator, whether they’re just starting out or sitting on a 100,000-strong subscriber base, will know the feeling of sitting down in front of your computer, and beginning to edit.

For some people — usually those who go on to become editors themselves — this is a comfortable work environment. They can pull together a captivating video like an artist painting a portrait. Each colour, each stroke of the brush, adds layer upon layer of polish until, finally, you end up with a masterpiece. For me, however, it was a nightmare.

I just wasn’t born to be an editor. I liked the creation part of content creation, actually recording the video, but when it came to stitching it together I either couldn’t summon up the motivation, or would become frustrated by my inability to edit. And being a one-man band with a small VR comedy channel at the time, I didn’t have the funds to search for an outside editor.

These days I’ve given up on YouTube, and on content creation in general, but I am fascinated by the new generation of editing tools which could save creators from suffering the same burnout that I did.

If you’re a fan of TikTok, you’ll most likely have heard of CapCut. It’s owned by the same company — ByteDance — and although you can upload its short-form content to any platform, it seems tailor-made for TikTok’s algorithm. And, though I can’t verify this, I’m sure that using CapCut as your editor gives you a small boost in viewership on the platform.

Descript, on the other hand, is an entirely different beast. If CapCut is the flashy new tool which everyon’e excited about, Descript is the trusty hammer which has been around for longer but isn’t quite as trendy. Think Ronan Keating in dungarees, for example.

In this article, I’ll look at the features of each in detail, including the ‘killer’ USP each tries to promote. I’ll grade each piece of software in three key areas: features, cost and ease-of-use, before adding up the scores.

And at the end of it all, I’ll give you my recommendation for the editing tool you should be using as a content creator in 2026 to speed up your workflow and avoid the same burnout which ended my own content creation journey earlier this year.

Ready? Let’s dive in.

CapCut: A TikToker’s Best Friend

Features

If your content is naturally geared towards short-form platforms, CapCut will quickly become your best friend. This would have been very useful for my own content. My VR comedy videos usually consisted of a series of loosely connected skits (often lasting around the one-minute mark) compiled together. So that content fits CapCut’s platform exceptionally well.

Where CapCut shines is in its effects library. It contains the most comprehensive resource of audio and visual effects I’ve seen. And while many of the most useful parts of the effects library are locked behind a subscription (more in the costs section below) you can still edit most of your short-form videos using the platform’s free tier.

Those effects are more than simple colour or lighting effects, too, they really can elevate your content and help you to catch the retention that’s so important to those clips going viral.

To give you a real-world example, I remember spending an hour on a Saturday morning creating a lighting effect in Adobe Premiere Pro. The effect was only to be used for one short section of my video, but I wanted it. And so I created it. And then I found CapCut and discovered that the same effect I’d laboured over was able to simply be dragged and dropped from the effects library onto my timeline.

If you want to turn your content into short, Hollywood-level clips, CapCut is the platform to go to.

As well as the effects, CapCut also contains text editors and automatic subtitles, as well as keyframe animtions and chroma keys.

The downside of CapCut is the fact that it’s very much a traditional editor. To go back to our artist analogy, CapCut gives you a far better brush, but you’re still having to do the painting yourself. And for creators with a to-do list a mile long, the process can be mentally draining.

FEATURES SCORE: 5/5 ⭐⭐⭐⭐⭐

Cost

CapCut’s free tier gives you limited access to the platform’s templates, effects, stickers, music and fonts. As I mentioned in the features section, what you get is usually enough for a smaller channel to build up an army of clips, but for larger channels, you might want the variety of the paid tier.

In addition, the free tier gives you 1GB of storage–which you’ll quickly use– and basic AI-powered tools like scripting with AI, and converting long-form videos into shorter clips.

By comparison, CapCut Pro gives you full access to that comprehensive effects library, as well as more than a 1.0TB of storage. You also get the more advanced AI features such as the ability to lip sync, create AI avatars, and create automatic captions from your videos.

As a creator in the sunset of their career, I happily paid for a monthly subscription to CapCut for the automatic captions alone, since this was the feature which eliminated the worst pain point for me: manually typing out captions.

CapCut’s Pro Plan costs $19.99 per month for individual users, or if you can afford to pay $179.99 up front, you can pay for a whole year.

And if you manage a team, plans for more collaborative work start from $24.99.

COST SCORE: 5/5 ⭐⭐⭐⭐

Ease of use

Since CapCut is available as an app and as a standalone piece of software on your PC, it’s incredibly accessible.

The mobile app is very easy to get to grips with, with clear layers on the timeline for your visual, audio and text elements. I speak from experience where even when you’re putting together a complicated clip involving multiple sources and layers, it’s easy to do. Hell, I managed to edit a vlog from a day out at London’s MCM ComicCon in 2024 on the train home.

The standalone app features the same layout, just on a larger scale. As someone used to the feature-heavy interface of Adobe Premiere Pro, I found CapCut’s somewhat simplified layout easier to get to grips with, but just as hard to master.

EASE OF USE SCORE: 4/5 ⭐⭐⭐⭐

Descript: The Architect Of Long-Term Success

Features

If CapCut is the toolbox which allows you to assemble your content more easily, Descript is the workshop which means you don’t need to assemble the content yourself any more. And for some creators, that will be a godsend.

What makes Descript work so well is its deep AI integration. You don’t even need to hunt for the tools you want — you can simply tell Descript to remove all of your retakes and awkward silences, or to find the viral clips from your long-form video, and it will just…do it.

The feature I loved most of all, though, was the text-based editing. I used to spend hours re-looking at my scripts to find the phrases that needed cutting, and then translate those cuts onto my finished video timeline. In Descript, deleting the relevant sentence from my script also deleted that portion of the video. It was, in a word, magical, and cut the time taken to do a rough cut from hours to minutes. And for a creator who is perpetually tired from 3am finishes, that’s a powerful tool.

Now, I pride myself on being quite an eloquent chap, but some of the creators I recorded my VR skits with weren’t quite so gifted, often needing re-takes and resorting to filler words (think ‘erm’, ‘uh’ and the like) when things got hectic. Descript could remove all of those in one click, leaving me with a better-quality video as a result.

Descript’s AI features go even further if you’re looking to clean up sound, because its Studio Sound feature doesn’t just clean your audio, it can make your cheap microphone sound like a professional setup. And if you end up stumbling or saying the wrong thing, Descript can synthesise your voice to cover the mistake. Think about that for a moment — making a mistake used to mean you’d need to re-record that section, whereas now, it can be covered with just a few clicks.

Finally, when your social battery is running low, Descript allows you to take a rest and not even look at your camera. You can read from your script, and the AI shifts your gaze to look directly into the camera lens.

Granted, Descript doesn’t offer many of the flashy and trendy effects which CapCut has, and for some creators, that will dampen its appeal, but for YouTubers and streamers looking to streamline their workflows, it’s worth a look.

So are there any downsides? Well, yes. Descript is mostly cloud based, so it can be sluggish when working with multiple 4K video tracks or complex B-Roll. And it also means that you need a good internet connection to access its most advanced features — there’s no offline AI heavy lifting here.

And as I’ve alluded to already, Descript lacks many of the flashier effects which have drawn so many creators to CapCut.

FEATURES SCORE: 4/5 ⭐⭐⭐⭐

Cost

Sadly, there’s no free tier with Descript. The entry-level option is called Hobbyist and costs $24, or $192 if you choose the annual plan. You get 10 media hours per month, along with 400 credits to use with Descript’s AI services. Speaking of which, you get the full suite of those, including the powerful filler word removal and AI speech tools I mentioned above. Plus you can export in 1080p without a watermark.

Moving up to the Creator tier will cost you $288 on the annnual plan, or $35 on the monthly plan. You get 30 media hours per month, 800 AI credits and you can export video in 4k.

As well as access to more advanced AI tools, you can generate video using the latest AI models, use footage from Descript’s stock media library, and get more media hours and AI credits if you need them.

It’s a punchier cost than CapCut for sure, but Descript’s AI tools are the more advanced set for creators looking to build long-term sustainability. And to put that price into context, the cost per month is about the same as the two pizzas you’d definitely need if you were manually editing your videos.

COST SCORE: 4/5 ⭐⭐⭐⭐

Ease of use

Let’s face facts here. Descript isn’t the flashiest program to use, and it’s not as intuitive as CapCut. Like CapCut, Descript offers a mobile app, but I’m willing to bet that most creators will use it from their computers.

Descript also isn’t as pick-up-and-play as CapCut, or indeed as most other editors. Text-based editing requires you to ‘unlearn’ what you thought you knew about conventional editing, because you won’t be looking at clips, but rather your scripts. But the upside to that is that editing your video by editing your script allows you to get the job done quicker.

And thanks to its powerful Underlord AI tool, you don’t need to know where the correct buttons are like you do in CapCut — you just need to know what you want the tool to do. And most of the time, it’ll do exactly what you want. For that reason, despite being the equivalent of Ronan Keating in dunagrees, Descript has all the star power of Boyzone.

EASE OF USE SCORE: 5/5 ⭐⭐⭐⭐⭐

CapCut vs Descript: The Verdict

Which of these tools you end up using will ultimately depend on what you want from your content. If you’re looking for a one-tap solution with the biggest chance for a single clip to go viral, CapCut’s intuitive layout and accessible pricing makes it the go-to choice. And as I mentioned at the top of this article, some of my best-performing TikToks came from using CapCut, although I have no evidence to suggest that there’s a subtle boost applied to such videos.

Note: I don’t even have an affiliate link for CapCut. I’m telling you to use it for your short-form edits because it’s the right tool for the job, not because I’m getting paid.

I’d urge creators of all sizes to consider Descript, though. Yes, it requires a bit of a rethink in terms of the way you edit your content. And yes, it is the more expensive choice. But this desktop-first powerhouse is the king of long-form editing for a very good reason. Its powerful AI tools can really help to take the sting out of your workflow, giving you quality control at scale.

Before I give my final recommendation, though, lets look at the side-by-side scores.

As someone who works infinitely better with words than with video, Descript’s script-editing process would have suited me naturally for producing my own content. It’s the one I would choose here.

Update: My deep dive into Descript (the ‘brain’ of my setup) is already live here. Check it out to see how I handle the long-form side of things

Note: This article contains affiliate links. If you click and buy, I may earn a small commission at no extra cost to you. I only recommend tools I wish I’d had during my 100k journey.

The clock on the side of my desk loudly blared that it was 3am. The room behind me was dark, quiet. The rest of the house deathly silent. Even the cat had finally given in and waddled off to bed. But for me, the night was far from over.

You see, I’d just hit publish on a new YouTube video. I’d spent all day recording, editing, researching the niche I was in, coming up with a hookable title and hashtags, and finally uploading the finished product. I was beat, but even though I was finally able to hit publish on the YouTube video, I was far from done.

No, now the real work began — and it’s the part of content creation I don’t miss the most. Full disclosure: I used to run a 100,000-subscriber YouTube channel built around VR comedy and GTA roleplay, and a 90,000-follower TikTok channel built around the same. I made videos, streamed on Twitch, and tried like hell to pursue the content creator’s dream. And it nearly broke me.

Quick summary: which should you choose?

For burned-out creators looking to make the most of their long-form content, OpusClip is the clear winner here. However, creators who thrive on hard analysis and doubling down on what their audiences want will find Munch’s blend of features very useful, despite the higher price.

These days, I game for fun, and while I’ll occasionally publish something to TikTok, it’ll be on my terms and not for views, but for the memories. Now I’m passing on what I learned — from my successes, but also my failures — so that my mistakes don’t need to be theirs.

Returning to that dark room at 3am, I remember feeling exhausted at the mere prospect of more editing, but there was no way to avoid it that I could see.

After all, the YouTube video I’d just posted was in a horizontal format, whereas Facebook videos, Instagram Reels, TikTok and YouTube shorts required a portrait-style video. For that I’d at the very least need to re-crop my existing video. But in order to give each clip the best chance of virality, I’d also need to add captions, extra sound and visual effects to keep audience retention as high as possible, and trending audio where neccessary.

The extra workload involved was, in a word, soul-crushing. And it often came at the end of what had already been a very busy day.

If I’d have used the new generation of AI-driven clippers, perhaps my content creation journey need not have ended. And in my research to find the best AI clipping software, two names kept cropping up — OpusClip and Munch.

In this article, I’ll look at the features of each in detail, including the ‘killer’ USP each tries to promote. I’ll grade each piece of software in three key areas: features, cost and ease-of-use, before adding up the scores.

And at the end of it all, I’ll give you my recommendation for the AI clipping platform you should be using as a content creator in 2026 to speed up your workflow and, hopefully, get to bed long before 3am.

Ready? Let’s dive in.

Before we start: how I rate and review products and services

For each product, I look at three key areas: features, cost and ease of use. Features encompasses the abilities of the product, including what it offers in comparison with rivals, and how each of those features improves the overall impact of the product. As a creator, I’ll specifically be looking for features which can enhance long or short-form content, and be easy to understand. I’ll pay special attention to AI-enhanced features, too, such as video-to-text generation and text-based editing.

In terms of costs, I’m looking for products which provide maximum return on a creator’s investment. I’ll set out what features you can expect to use in each tier, and note if there is a paid tier available. I’ll also give my recommendation on which paid tier of a product is worth going for, if any.

Ease of use is all about maximising your productivity. I’ll be looking for features which are easy to understand, and a user interface which makes it easy to do what you’re trying to acheive. For this, products should offer clear instructions and icons, and come with documentation or tutorials to help customers get the best out of them.

OpusClip: the efficiency king?

Features

OpusClip looks like an all-in-one AI solution, with tools designed to avoid the 3am burnout I mentioned earlier. For one thing, it has the feature which would have saved my bacon back in the day — an automatic tool to find the most sharable, viral moments in your long-form content and convert them into short-form clips.

And it goes one-step further, applying automatic captions to those clips, saving you even more time.

I tried this feature and, frankly, I was gobsmacked. Uploading one of my older YouTube videos — you only have to provide the URL, rather than downloading the video itself — started the process. I used 8 credits to scrape that long-form video into shorts, with the whole process taking around six minutes.

From my one 8-minute YouTube video, the software picked out 16 viral moments, creating the short-form clips for each one and assigning it a ‘virality score’ based on how well it was likely to do on short-form platforms.

From there, I could post to my social media channels directly (no more logging into each platform to post separately), or download the videos to my PC for storage. Plus, I could edit each clip if it wasn’t exactly what I wanted.

I like that I can further refine the created clips directly within OpusClip’s interface, adding B-roll, changing the prompt for the AI clipper or enhancing speech (something which I needed at the time, since my microphone back then was, in a word, terrible).

What stops the AI clipper getting five stars in my book is that, while it generated 16 short clips, not all of them were bangers. A couple in there were either just not funny, or too short to gain any traction. A good example of this was a 16-second clip which was more an off-the-cuff comment rather than a skit for the video.

And while it’s great to have automatic captions on the clips, a lot of the templates are quite samey.

FEATURES SCORE: ⭐⭐⭐⭐(4/5)

Cost

Free trial? 90 free credits, valid for 7 days. Beyond that, you won’t be charged unless you choose to upgrade. Full range of features except fast queue.

Starter plan €15/month, Pro plan €29/month (billed at €174 annually), Business plan custom price

OpusClip’s free trial credits should last solo creators about a week or so, which is enough to get to know the platform and its capabilities. I like that you aren’t charged to upgrade after your seven-day free trial is over unless you choose to do so, which avoids you automatically signing up to a subscription that you won’t use.

The Starter plan ($15/month) gives you 150 credits, the all-important AI clipping feature, animated captions in 20 languages, and the ability to auto-post those short-form clips to YouTube Shorts, TikTok and Instagram, as well as download them.

For YouTuber teams or channels with more content, the Pro plan ($29/month, but billed annually) gets you 3600 credits per year, with two people able to access your workspace within the platform. You can pay extra to upgrade to four slots for your workspace.

Additional features with the Pro plan include AI B-roll footage — I could have made good use of that in my videos — as well as the ability to export directly to editing suites like Adobe Premiere Pro and DaVinci Resolve. You can also select different aspect ratios, and schedule your clips with an integrated calendar.

Business users — by which I’m taking to mean large YouTuber teams running multiple channels– will need to contact OpusClip to discuss cost, but you get everything in the Pro tier, plus dedicated storage, a custom API, dedicated support and enterprise-level security, so nobody has access to your content.

For most creators, I’d be tempted to stick with the Starter plan, which should significantly reduce the time taken to do your short-form clips.

COST SCORE: ⭐⭐⭐⭐⭐(5/5)

Ease of use

Since everything is done within the OpusClip platform, it’s easy to get along with. I like the dark aesthetic (no bright colours at 3am, thank you) and that I could see all of my clips in one place.

And while there are plenty of features to get to grips with, everything is clearly labelled and presented, so the platform doesn’t feel as daunting to use as some other AI clippers out there.

I did find that the transcript of my clips was a bit small to read on a larger screen, so I had to lean in close to fine-tune the transcripts of my clips.

EASE OF USE SCORE: ⭐⭐⭐⭐(4/5)

Munch Studio: the context specialist

Features

If OpusClip is more geared towards the one-person creator, Munch (or Munch Studio, to give it its proper name) seems geared towards teams. The flashy front page of the website offers to run your social media in 10 minutes a week. It’s a big claim, but then again, Munch offers some big features.

Just like with OpusClip, I could upload my long-form content to then be cut down using the AI clipping software.

What Munch does over OpusClip is provide a more well-rounded content strategy. As well as cutting down your video for shorts, it can also draft posts for you to publish on X, Instagram and even LinkedIn. In short, it’s more of a corporate face than one geared exclusively to creators.

I like that if you do want to make use of those posts, you can use Munch’s smart schedule feature to ensure those posts hit at the best times for audience engagement.

And if you don’t want to schedule those posts, you can save them for later in a dedicated tab.

When it came to actually creating clips, Munch includes analysis tools to try and tailor your AI-created clips to tap into TikTok’s algorithm, making them pop that much more.

That trend mapping is incredibly clever — the AI tells you why it picked the clip it did. That means if the clip does well, you can try to recreate and expand on it for future clips.

I also liked that the subtitles used were far more customisable than in OpusClip, allowing me to use a template which viewers wouldn’t see on 100 other channels. For niche content, such as my GTA RP streams, that would have been ideal.

FEATURES SCORE: ⭐⭐⭐⭐(5/5)

Cost

As standard, you get a seven-day free trial, which then converts to a paid tier with an annual cost of $456, or a monthly cost of $48. That’s punchy, to say the least, being more than double the annual cost of OpusClip.

I dislike that, unlike OpusClip, Munch’s free trial automatically converts to a paid subscription, though Munch says it will email you 48 hours before your trial ends.

There’s no tiered solution in Munch, with everything thrown into the single paid tier. That includes the following:

• AI clipping
• Built-in video editor
• Create posts from websites and blogs
• 500 minutes/month video repurposing
• 100GB of video storage
• Download to you PC
• Personalised content strategy
• Auto-publishing to all social patforms

That monthly cost is very high, but then again, Munch’s ability to analyse your content and explain why it might perform well is powerful.

In my GTA RP days, it would have helped me to see whether my latest chase with the Police, or a deep emotional conversation, was picked up better by my audience. That trend mapping gives you an insurance policy against producing content that nobody is searching for.

COST SCORE: ⭐⭐(2/5)

Ease of use

Munch’s interface is a lot lighter by default (not good for that 3am glare) and I’d say the system requires a bit more of a learning curve. I had to look up several YouTube tutorials before I was comfortable with the interface.

I like that the buttons to create your content are of a good size, though, making them easy to find.

Munch’s suggestions for trending hashtags and viral hooks looked to be bang on trend, too.

Plus, Munch gives you the ability to add royalty free music from its extensive library, which for smaller creators without access to external sound libraries could be a game-changer.

EASE OF USE SCORE: ⭐⭐⭐(3/5)

OpusClip vs Munch: the verdict

OpusClip wins the features battle for me. While Munch is better at picking out the best moments from your long-form content, OpusClip creates more clips and seems to give you better control over them.

And then there’s the cost. I would find it hard to recommend Munch to single-person creator teams. And in fairness, Munch itself seems to be marketing itself more towards larger channels who have teams of editors.#

Before I give my final recommendation, though, lets look at the side-by-side scores.

OpusClip: pros and cons

PROS

• Speed in turning around content
• AI model makes it easy to keep the camera on your main character
• Generous free tier

CONS

• Rigid Credit System
• AI model can miss crucial context
• Captions and B-roll styles can feel samey

Munch: pros and cons

PROS

• Munch tells you why a clip is being made based on what is currently viral.
• Powerful tools write the SEO-optimized captions, hashtags, and even social posts for you.
• Subtitle editor and branding tools are far more advanced than Opus

CONS

• No mobile app
• Steeper learning curve than Opus
• Hefty investment for smaller creators

CONCLUSION

OpusClip’s blend of features for a very accessible price make it the perfect choice for the burned-out creator who wants to get to bed at a reasonable time. While Munch is the better fit for creators who want to dominate with data-driven analysis, its price is too steep to recommend for most people.

Update: My deep dive into Descript (the ‘brain’ of my setup) is already live here. Check it out to see how I handle the long-form side of things

Note: This article contains affiliate links. If you click and buy, I may earn a small commission at no extra cost to you. I only recommend tools I wish I’d had during my 100k journey.

I remember the moment I hit 100,000 subscribers on YouTube.

It was Boxing Day, just a few months after YouTube had introduced its Shorts feature. Channels were blowing up because of it, and I’d had a couple of my own shorts picked up through the feed. Those Shorts weren’t anything special, but they hit upon some sort of virality, and the numbers had been spiking for days.

I was like a junkie staring at that analytics window. I could watch the sub count on my channel rise higher and higher. First to 10,000 subs, then 20,000, then 40,000. It was insane. A friend at the time made me a custom plaque to celebrate hitting 10,000 subscribers, but by the time it arrived through the post I was actually celebrating hitting 50,000.

And sitting there, late on that Boxing Day evening, with the magical 100,000 figure edging closer, I felt excited, hopeful and utterly exhausted.

You see, like most of the roughly 60 million YouTube channels which actively upload content, I was a one-man band. I was the face of the channel, the ideator, the editor and the social media expert. I was trying to do four jobs which are essential to running a successful channel, on top of my day job. And looking back, I was doing none of them very well.

In the months following hitting 100,000 subscribers, I pivoted my content so many times that I’m amazed anyone stuck around. I was trying to keep the channel going while reducing my workload, and the result was sub-par content. In a few short months, my audience had largely moved on.

The same thing happened on TikTok. I streamed the free battle royale game Fortnite on weekend mornings. My audience had to follow to play, and it helped to quickly catapult my account from around 12,000 followers to 110,000 followers. Again, for a few moments I felt on top of the world, before the crushing tiredness set in and I knew I wouldn’t be able to sustain it.

So at the end of 2025, I walked away. These days I game for myself, upload the odd clip if there’s something of note which happens, but nothing beyond that. I feel like I’ve regained sanity in the process, too.

But here’s the thing — had I used the right tools and the right workflow, I need not have walked away at all.

The truth is that being a content creator has never been harder, with more competition and more platforms to be present on than ever before. But thanks to the plethora of AI tools which are now available, in some ways getting your content seen has never been easier.

And that’s what I plan to showcase here on Medium. I’ll be showcasing the kit I used to accelerate my content creation growth, and also the AI workflows I’ve since discovered which could help other creators avoid burnout.

In short, my fate doesn’t have to be yours.

Introducing the AI team for content creators

The days of being a one-man creator are over, because now you can have an AI team at your fingertips.

I’m sure every creator can empathise with the dread I felt looking at my editing software and a huge timeline which needed editing not only into a YouTube video, but then into social media clips.

And it wasn’t only cutting up the footage. I also needed to add captions, sound and visual effects, funny green screen footage and end-screens. It was exhausting to even look at.

So if I was still going today, here’s how I would incorporate AI into my creator workflow, and how it would save me time — and energy — in the process.

The video editor — Descript

Descript is an AI-powered video and audio editor which allows you to edit a long video simply by taking out words from within it. So I wouldn’t be staring at a video timeline for six hours at a time, but rather deleting words from a script, much like writing this post.

As well as helping you edit, Descript also provides powerful AI voiceovers, perfect for the new wave of faceless channels.

I’m currently updating my partner link for Descript, but you can find it and the rest of my burnout-proof streamer recommendations on my curated list here.

The content splitter — OpusClip

Once I’d made my long-form video and published it to YouTube, along with the catchy title and thumbnail, SEO-optimised description and captions, I would love to think I’d be done. But no, I’d then need to re-open my editing software and re-package my one long video into 10 shorter clips to put out on TikTok, Intagram and YouTube Shorts.

OpusClip would have saved that pain. Those viral shorts which helped me to hit 100,000 subsribers would have been sorted in just a few minutes, because the platform combs through your video (or your stream VOD), highlights the most sharable moments and automatically creates clips from them.

The content scaler — Repurpose.io

When I was in the middle of the content grind, I used to hate having to upload my short-form clip onto several platforms. I’d log in to TikTok, post, then remember that I should do Instagram as well, so try to log into that having inevitably forgotten my password. And after I’d done that there was still Facebook, X and YouTube Shorts to consider. I was drowning in terms of content distribution.

Which is where Repurpose.io comes in. I’m genuinely annoyed that I didn’t know about this platform sooner, because it’s been designed to remove the very pain point I’ve described above, by automatically seeding your content to every platform you connect.

The SEO expert — VidIQ

I mentioned earlier on how I pivoted my content following hitting 100,000 subscribers. It was a totally stupid move, but I didn’t know any better because, again, trying to run a successful channel all by myself meant I was doing every job badly, instead of one or two jobs well.

What I should have done was muscle down on the niche my audience was watching, and try to predict what they would want to watch next. Instead, I was throwing my content at the YouTube wall and hoping some of it would stick.

VidIQ can stop you from making the same mistake I did. I was guessing what my audience wanted to see, but you don’t have to. VidIQ can give you actionable insights like titles, hashtags, descriptions, thumbnails and all of the back-end elements which help to maximise the reach of your content.

UPDATE MAY 13: I’ve just published my comparison of ElevenLabs and HeyGen. If you’re struggling with the ‘presenter’ side of the 100k grind, start there.

Note: This article contains affiliate links. If you click and buy, I may earn a small commission at no extra cost to you. I only recommend tools I wish I’d had during my 100k journey.

“Is this the cost of progress? We’ve solved the problem of having to think.”

That was what Advait Sarkar said during a recent TED Talk forwarded to me by a colleague.

His point is simple; by integrating AI into our workflows, we’re in danger of outsourcing critical thinking.

Because we’ve all seen what happens when we’re not challenged any more. AI tools, in their quest to be helpful to the user, are inclined to agree with us rather than challenge us. They obey rather than tell us there’s a better path.

When I’m interacting with any number of the Gemini Gems which I use on a daily basis, I remember that part of their instructions are to challenge me; to tell me if what I’m suggesting isn’t the best course of action.

During his talk, which I recommend you watch, Sarkar said that we must act now to steer the wave of generative AI tools which stand ready to transform our working lives towards humanistic values. They must be told not just to obey commands, but also to question them.

An army of AI tools which agree with us 100% of the time may lead to increased productivity and a reduced workload for businesses, but the cost may also be a dumbing down of the critical thinking which drives business in the first place.

AI can automate the known and provide faster work. But what we need AI to do is ask the right questions and explore the unknown — that is the path to higher quality work.

What we need to appreciate if we want to use AI to keep our critical thinking actually goes against what most businesses are looking to use AI for — that to keep our expertise, we need to sacrifice pace. In other words, we need to be okay with a high-quality AI-enhanced answer to a problem being slower than the alternative.

As part of his talk, Sarkar showed a working prototype model built by him and his Tools for Thought team at Microsoft Research in Cambridge. In a fictitious example, the model was assisting a worker with digesting an industry report.

Rather than just outlining the document as a summary, the model offered its own ‘provocations’ to assist in the worker’s thinking. For example, part of the example document mentioned sustainable packaging, at which point the model pulled out a note that the worker’s brand could incorporate such packaging.

The result is that the worker is not only given a summary of what’s presented to the model, but is also prompted to keep thinking — to enhance their work having been stimulated by the provocations.

And even when the worker chose not to act on one of the provocations presented by the AI model, that wasn’t deemed as failure — because the very act of deciding against a provocation is part of the critical thinking the model is intended to enhance.

One of the most powerful uses of this model came when the worker was writing for themselves. The model wasn’t trying to auto-complete their sentences, but rather raise questions for the worker to answer themselves. And not just questions, but counter-arguments and fallacies. And there wasn’t a chat function in sight — the worker was not talking with an AI, but being assisted by one quietly and efficiently.

And crucially, the end output was high quality and undoubtedly human — became it was never written by AI, just assisted by one.

It’s exactly this thinking which I have tried to adopt when creating the Gemini Gems me and my editorial team uses. Take our sub-editing Gem, for example. It’s not designed to merely point out grammatical or spelling errors in supplied copy, but also to provocate the writer by asking questions, and by prompting them to add the kind of context which turns a regular article into a piece of expert content.

Sarkar ended his talk with a single thought: what would we rather have, a tool which thinks for us, or a tool which makes us think?

I know which I’d rather have.

In an episode of the cult sitcom The Office, Steve Carrell’s hapless manager Michael Scott loudly declares ‘bankruptcy’ by shouting it in front of his confused colleagues — believing that by doing so, his financial worries will be solved. The moment has become a centre point of internet memes since, but also parallels the way in which a lot of brands are thinking about their AI-assisted content.

In academic texts, students are required to disclose their use of LLMs, whether used to research, draft or edit. And the same is true of scientific journals. This makes total sense, given the weight of the articles being written and the implications of their content. But for publishers, who are not required to disclose their AI content (yet, at least), the question remains: should they? And, if so, how should that look and what AI-assisted threshold needs to be passed for it to apply?

Working for one of the UK’s largest publishers, this has been a topic of great debate internally. Our initial thoughts arrived at a one-size-fits-all framework where, since we wouldn’t be publishing anything solely written or created using AI, we wouldn’t disclose its use.

But over the past year, standards have shifted, as have the tools available to us and the quality of output that they’re able to achieve. So when I was asked to reflect on the AI editorial principles we use as a publisher, a question posed itself: did that approach need to change?

In my mind, yes, it did.

Where AI has been solely used to create content, publishers should declare it. But what about where it has been used to assist in creating the content? Or in cases where the AI has written the draft, but that output has been verified and expanded upon by a human writer, should the same apply?

My thinking led to two approaches. And here, we’ll explore both before I tell you where I ended up – and what I recommended to my colleagues.

Route 1: Don’t disclose AI content, ever

One side of the argument around AI content disclosure says that AI is merely a more advanced form of spell-checking. It checks tone, resonance and style, but so long as it’s not adding to your content, there’s no need to disclose its use. The work being presented is still fundamentally human.

The stats show this is already the case. According to a recent study, 83% of advertising executives admit to using AI at some point during the creative process, but less than half of that figure disclosed the use of AI content in their final pitches.

So, in the same way that we don’t disclose using a dictionary, a calculator or Google Search to aid in our creative process, is there any need to disclose the use of AI?

The benefit of not disclosing content is clear for businesses and brands. A global study by advertising firm KPMG found that less than half of consumers (46%) were willing to trust AI systems. Of those, just 16% said they strongly trusted AI-generated content.

Then there’s the terminology used to associate AI content. You’ve probably seen the term ‘AI slop’ referred to online and on social media. And even when that’s swapped for the ‘contains AI-generated content’ labels used by Facebook, Instagram and TikTok, most users still associate the label with low-effort, low-quality content. To label that content yourself, then, could mean making it inherently less appealing to your audience.

And if you do choose to label all of your content as being AI-generated, even if AI has only been used to assist in its creation, the label you once sought to make clear would become meaningless.

The stats suggest that’s already happening, with one survey revealing that 48% of respondents distrust the accuracy of AI labels used on social media. Less than one-in-five (19%) said they would trust such labels.

One final point here is that under current UK legislation, it is not mandatory to label AI-generated content. And while the Government’s consultation on labelling such content — which ended in 2024 — ultimately deemed it to be beneficial for readers, it also noted that there would be ‘significant technical challenges’ involved in doing so.

That’s one side of the AI content disclosure argument — now, let’s see what the other extreme looks like. And after that I’ll reveal where my thinking ultimately ended up.

Route 2: Do disclose AI content, all the time

While the UK doesn’t have any legislation mandating the labelling of AI content, the European Union (EU) shortly will do. The AI Act will come into force in around August of this year, and will force publishers and businesses to include clear labelling for AI-generated content. There are big penalties for non-compliance. And while the UK has said that it will not formally follow the AI Act, businesses showcasing content in Europe will need to.

So, it’s in the interest of publishers to declare their AI content now, before they’re mandated to do so. This might sound like radical transparency, of the kind which signals fear of reprisal rather than innovation, but by doing so brands could future-proof themselves against further legislation.

For readers, labelling AI-generated content could help to build trust with brands. Rather than feeling mislead, or that publishers are being ‘dishonest’ when serving AI content to readers, labelling content signals transparency in how that content was made, allowing users to make informed decisions about what they’re reading.

In an age of dishonesty and distrust — particularly when it comes to the media — labelling any AI-assisted and AI-generated content could signal that you, as a brand, have nothing to hide.

The problem with this approach is in how much protection the AI-generated content label actually affords you. The evidence suggests that the answer is ‘not much’. Publishers are still responsible for what they publish, even if that content originated in a machine, rather than a human mind. As Michael Scott found out, shouting ‘this is AI content’ at the top of your lungs doesn’t provide a magical fix or get-out-of-jail-free card in the event of a brand-threatening hallucination or libelous statement. You can’t disclose your way out of a lawsuit.

Your writers may also not be fans of this approach. If half of the articles on your website are labelled as being AI-generated, it could make your brand’s content look as if it rolled off an assembly line, and de-value the expertise of your remaining human staff. And in 2026’s SEO landscape, expertise is one of the few signals Google is still listening to in order to push content.

So, there’s our second radical move, and the complete opposite of the first approach. But which is correct, and which did I ultimately recommend that my brand and wider business adopt?

What I suggested: the threshold strategy

It won’t surprise you to learn that where my thinking up was in neither of the two camps mentioned above. Instead, I recommended a threshold strategy, where we would disclose content as being AI-assisted where AI was significantly used as part of its creation.

To make this approach work, we needed to identify what ‘significant’ meant. After all, if the content had been run through a Gemini Gem designed to sub-edit out any spelling or grammatical errors, I would argue that it doesn’t need to be labelled as AI-assisted content. But where AI has been used to write the content, and that content has not been further edited before publication, it should be.

But writers, editors and publishers require less nuance. What one person agues is significant might not be in another person’s definition. We needed a number.

I arrived at the following: if the finished article was flagged as being more than 50% AI-generated using a third-party AI content detector such as ZeroGPT, then that content should be flagged for further editing by a human.

In practical terms, where more detail might be needed, I proposed the following three-tier strategy for AI content disclosure.

• Level 1: AI used for research, transcription, headline suggestions or SEO metadata. No disclosure needed.
• Level 2: AI used for content briefs, H1/H2 structures. No disclosure needed, provided the published content is 100% human-written. If not, refer to Level 3.
• Level 3: AI used to write the first draft, but a human polished it. Disclosure required if more than 50% AI-generated content is detected. Article flagged for additional editing by a human.

You’ll notice that the common factor in all three levels is the so-called ‘Human in the Loop’ — the rule that no content completely generated by AI will be published without human editing. And even when AI is used to create content, it is signed off by a human, before disclosure is even considered.

Why “Human-in-the-Loop” is my recommendation

In my opinion, and contrary to the Michael Scott approach, I don’t think that simply shouting ‘AI-generated’ content is enough, even when that label is warranted. AI use cannot simply be labelled, it must be governed.

Businesses of all shapes and sizes should decide their own approach to AI, but for my industry, the bottom line has to be this: the human must always be in the loop.

What do you think? Let me know in the comments below.

I write about AI from a writer’s perspective on my free Substack. To sign up and receive my latest updates, click here.

You’ve probably experienced this all too frequently — you’re being asked to do more with less, cramming more productivity into a day which, from the outset, isn’t exactly overflowing with downtime. Sprinting from day to day has become the new normal, and without something changing, I fear could lead to huge burnout.

The signs are that this is already starting. Mental Health UK reports that 91% of workers reported periods of high stress in the past year, with one-in-five working adults taking time off over the course of the year due to poor mental health caused by stress. Most worryingly, the charity reports that two-thirds of the UK workforce are showing symptoms of burnout.

As a writer working in a fast-paced, deadline-heavy environment, I’m not immune from this. And, speaking candidly, poor mental health has plagued me in the past. I can remember full-on panic attacks and breakdowns at the thought of my workload. So I was one of the first to embrace the productivity gains brought by AI. And it has done what I once thought impossible — brought me room to breathe again.

Here, I thought I’d share some real-world examples of how I use AI in my role working for one of the UK’s biggest publishers. Perhaps some of these tools or workflows can help you too.

The TL;DR: I don’t use AI to write for me; I use it to handle the cognitive load of research, SEO logistics, and initial sub-editing so I can focus on the “human” part of the story.

NotebookLM: Listen While You Work

I receive large reams of documentation fairly regularly on a variety of subjects, and usually accompanied by pictures which need re-sizing and editing.

So, while I tackle those — a process which might take twenty minutes – I upload the supporting documentation into NotebookLM and ask it to generate a podcast using the materials. Before starting, I change from the default setting to ‘long’ and update the instructions to tell the podcast hosts to approach the subject with the practiced cynicism of a seasoned journalist.

The result is often a good starting point for the topic, meaning that when the time comes to give the material a deeper read, I’m already somewhat informed. In short, AI gets me to the point where I can be the expert I’m paid to be, quicker.

Gemini Gems: My Content Advisor

When it’s time to put proverbial pen to paper, modern writing for online audiences requires more than just tapping away — it requires SEO guidance, knowledge of H1, H2 and H3 headings, keyword optimisation and (new for 2026) the answering of the kind of questions likely to be surfaced as part of Google’s own AI Overviews (AIOs).

Incidentally, those AIOs are an increasingly important part of content optimisation, since the value of traditional SEO is being quickly eroded. Google AIOs are now being shown at the top of your search results for 55% of enquiries — pushing the top search result further down page one, and some entries which would traditionally be on page one onto page two. And as you’ll likely know, anything other than page one might as well be nowhere.

All of that means that before writing, I need a content brief. And fortunately, that’s exactly the kind of task which a Gemini Gem is well suited for.

To set this up, I gave the Gem a series of instructions including how I wanted the content brief to be laid out, the information to include and the suggestions of keywords I should be targetting. I also uploaded our brand style guide, our latest SEO guidance and several examples of ‘good’ content briefs on previous topics.

The resulting content brief is easy to understand and quick to action — whether for myself or any freelance writers I might comission for the content. Using a Gem for this task has allowed me to deploy content briefs at scale, giving writers a blueprint for their articles. I don’t see this approach as diminishing creativity, but instead guiding writers on how to get their content seen by as many eyes as possible.

I can remember full-on panic attacks and breakdowns at the thought of my workload. So I was one of the first to embrace the productivity gains brought by AI.

My Own Personal Sub-Editor

In my opinion, all writing should be read back at least once before being published, whether by the original author or, in the kind of publishing environment I work in, by a sub-editor. Subs don’t get enough recognition in my book — they tighten up copy, suggest the kind of headlines which get readers clicking, and generally make my writing appear a lot better than it deserves to.

But budgets are shrinking across the board, while workloads are rising. And in my case, that means that the online content appearing on our website doesn’t get a dedicated sub. That’s beginning to sound like a problem a second Gemini Gem could fix, no?

And that’s exactly what I did. A second Gemini Gem was trained on multiple examples of ‘good’ writing for my brand, as well as our style guide and SEO guidance. It was instructed to provide no-nonsense feedback to writers, never outright demanding changes to their provided copy, but rather suggesting with reasoning. The result has been profound, both in terms of the quality of the copy, and the time saved in reading multiple stories two, three or maybe four times.

Now, the senior staff who still do a human pass on all of our written content face an easier life, because the original author has already run their copy through the gem.

And what of content security? If we were using a free tier of an AI tool to accomplish these tasks, I’d be concerned that all of our writing would be used to train another LLM. But with Gemini’s Enterprise tier, we have Google’s own assurances — for whatever they’re worth — that our content is not used to train its models. Anything we upload to Gemini, as well as NotebookLM, exists in a ‘walled garden’ of content. It is ours.

With Gemini’s Enterprise tier, we have Google’s own assurances — for whatever they’re worth — that our content is not used to train its models.

Becoming an AI-Assisted Writer

Incorporating these tools into my daily editorial workflow has been nothing short of transformational for me. For the first time in months — possibly years — I have room to breathe, not because I’m doing less, but because AI has removed some of the drudgery from my role. And I do believe that my writing has improved as a result.

Hopefully, seeing my own workflow might inspire others to adapt and adopt elements into their own. I’d be very interested to hear from others in the comments.

I delve more into this and other topics on my free Substack. To sign up and receive my latest updates, click here.

In early 2023, my workplace was set ablaze by the notion of using ChatGPT to supercharge our productivity. In short order, working groups were set up across job families. Editorial types looked at writing tasks, marketing managers started generating email copy with the hope of driving sales, and developers began to explore how we might integrate the whole lot into our existing systems. The buzz was practically electric.

I’d imagine plenty of businesses can report a similar situation. And as more AI tools came on stream — seemingly very quickly — we categorised, explored and tested them all. The result was a robust list of approved AI tools for the workforce to use, along with a policy governing their usage.

The effect of adopting AI tools as part of our workflow has been profound. The London School of Economics reports that employees using AI tools save the equivalent of a full working day per week, boosting individual productivity by around £14,000 per year. Twenty minutes saved here, ten minutes there — the productivity savings quickly add up to pounds and pence.

From cool tools to pain removal: the day I changed my training strategy

But now, something is changing. Those early adopters can no longer afford to explore new tools simply because they are new. Instead, they’re being asked to match AI tools to specific outcomes. In other words, AI now has to make a return on investment. For the early adopters, we’ve moved from generating curiosity to hard accountability.

It’s understandable why this shift has occurred. Employees could spend vital business hours exploring AI tools which might not have anything to do with their roles, simply because it falls under the banner of ‘AI’. And yet, the last thing I’d want to do is constrain creativity. That freedom to explore has so often given birth to new ideas and, ultimately, potential sources of new revenue.

But AI adoption across a business must meet the needs of that business.

In my role as a leading voice for editorial AI use for a major UK publishing house, I’ve been pushing not only for AI to be matched to specific KPIs, but to solve specific problems. Because I believe that when we can use AI to remove pain points for the workforce, we unlock its true potential.

I’ll give you an example. Our early rounds of AI training focussed on showcasing the tools and what they could do. And sure, the early adopters and keenly interested staff members took them away and began to use them. But most people didn’t, because we weren’t removing pain points in an already busy schedule.

So for our next round of training, I tried something new. I gathered colleagues into a room and asked them to explain the most tiresome, gruelling and downright unpleasant part of their day. And then I showed them how AI could help to take the load off. The result? A happier workforce, and AI adoption increased at pace.

Which KPIs to use?

Businesses might struggle to see which Key Performance Indicators (KPIs) would work best to evaluate the effectiveness of an AI system or platform, beyond the time saved. In my opinion, it comes down to three points:

1. Efficiency — how easy is the system to use both for AI-fluent colleages, and those who are new to AI?
2. Accuracy — is the output from the AI accurate, both factually and, for editorial purposes, does it follow the correct tone of voice?
3. Business lift — has the use of that AI system improved productivity, shifted conversion rates or driven sales?

By using these metrics, businesses could easily track the effectiveness of the AI platforms they’re evaluating, and pick the ones which best suit their needs.

Of course, picking the right suite of AI tools is itself a difficult process, because you don’t want to be tied into a lengthy contract with several different providers, while a one-size-fits-all solution launches a month later. But that’s a subject for another day.

Welcome, evangelists

But what of our gallant early adopters? The people whose energy and enthusiasm kick-started many business’ interest in AI in the first place? Well, they’re still sorely needed. Those people are your evangelists, your product soothsayers, your tech tzars. They are the people who will keep bringing new tools to the table, and should be encouraged to keep exploring.

And continuing to be an early adopter can yield big rewards. Reuters reports that 53% of people who consider themselves to be AI early adopters experience improved time management. Such people are, in my experience, often better at innovating and adapting to changing market conditions.

So continue to go boldly, to search and explore the widening pool of AI tools are your disposal. Continue to catalogue, to test and to report. And even if you have to do it within the constraints of business-specific KPIs, keep being curious. Because now, as always, the early adopters are critical on the road to an AI-enhanced future.

I delve more into this and other topics on my free Substack. To sign up and receive my latest updates, click here.

For writers and content editors, the rise of AI can look like a complete disaster. Think the asteroid hurtling towards Earth in Armageddon, or the, erm, asteroid heading towards Earth in Deep Impact.

What was once limited to a fringe technological tool is now everywhere. Whether you’re researching a purchase, making a reservation or looking for advice, the chances are that you’re interacting with AI long before talking to a human. And with that comes the slop — the predictable, LLM-generated text which some companies have decided to deploy wholesale rather than paying a writer to do what they do best.

AI slop isn’t writing, it’s mathematics in action.

The grammatically perfect but contextually void output of most LLMs is simply the logic behind the AI trying to piece together a coherent sentence. Because LLMs simply look for the next most logical word in the sentence, their output, though technically correct, reads with all the excitement of watching paint dry.

The damage done by AI slop is hard to quantify, but it boils down to this: in 2026, getting your content in front of readers is harder than ever, because they’ll need to wade through AI slop in order to get there first.

And readers are beginning to notice. They’ve seen the synthesised fluff being presented within Google’s AI Overviews or on individual websites. They’ve scrolled through the beige-baked TikTok videos hooked onto a viral search term, and they’ve become increasingly wary about clicking through to what could be expert content for fear of hanging over their details to yet another bot.

Here’s the kicker: that represents a real opportunity for writers.

Why? Well, AI slop is now everywhere, which means most writing served up by search engines is beginning to sound identical. Tone of voice is absent from a growing number of corporate texts, and readers are beginning to tune it out. When everyone has access to ‘good enough’ content, ‘good enough’ no longer serves its purpose. And that’s where writers still have a crucial role to play.

Let me be clear: AI-assisted writing can be a real boost for businesses. But while LLMs can write, they can’t add the context, analysis and expertise that your audiences have come to expect. They can give you the facts, but they can’t give you the human edge which might ultimately convert a reader into a customer. And that’s why I’m a huge advocate for the Human-in-the-Loop editorial framework.

Wait, what’s that, you say? Human-in-the-Loop is a way of thinking when it comes to editorial AI, that no matter what you’re using AI for, a human is an integral part of the process.

In practice for me as a writer, that might mean that I use AI to generate a headline, a content brief, an SEO guidance sheet, or even spark creativity for my introduction — but it never completes the job.

Here’s how it breaks down:

• The Drudgery (The AI): Headline variants, content briefs, SEO keyword clusters, structural skeletons.
• The Excitement (The Human): Contextual analysis, nuanced opinion, personal experience which converts readers to customers.

The benefit to me as a writer is simple: AI helps me to get to the point where I put pen to paper, quicker. By taking away some of the drudgery associated with what I’m planning to write, I can get on with the exciting part of writing (the writing) quicker. And that means my productivity increases.

To exist alongside and rise above the increasing tide of AI slop in 2026, writers need to consider themselves as more than that title — they need to be architects as well. Using AI for the heavy lifting of your content (think the research, structure, SEO) reduces time spent before you start typing, but the voice in the end product is clearly your own.

In short, then, AI can write the recipe, but you’re the chef.

And the resulting meal is one which, I think, brands will increasingly want to pay for. With AI slop dominating search engines, brands will pivot to expertise to help them rise above the tide. So writers who can utilise AI to assist, rather than replace, their writing skill stand to profit.

Being an AI-assisted writer is a badge which I don’t expect many to wear with pride. I think plenty of writers will view any kind of AI assistance as intruding on their turf, and plenty of that resentment is born of fear. But working in AI governance, I can tell you that most of that fear, at least in my experience, is unfounded. While AI has been adopted at my place of work, our thinking has always been about how to enhance our workforce productivity rather than replacing any person with an AI system.

And when it comes to commissioning my own content, I ask one question before using AI for the job: am I taking potential paid work away from someone else? If the answer is yes, I think again. If the answer is that without using AI, the job will fall back on me in any case, I crack on.

So should you, as a writer, a content editor or an editorial director be using AI in 2026? I think the answer is yes, but not to do the whole job for you. Think of AI as being like your research assistant, feeding you all of the information needed to allow you to work in peace, unburdened by logistics. You remain the human expert, even amid a sea of AI slop. And as readers become more attuned to glossing over AI content, that expertise is going to become a valuable commodity.

I delve more into this and other topics on my free Substack. To sign up and receive my latest updates, click here.