Have you ever watched a movie where robots and superintelligent AI aim to take over the world? While it might seem like pure science fiction, some experts believe a similar scenario could unfold in reality. Visionaries were increasingly convinced that as we edge closer to achieving artificial general intelligence (AGI), the importance of AI alignment grows exponentially.

In 2023, leading figures in AI research called for a pause in AI development, signing an open letter that read: “Powerful AI systems should be developed only once we are confident that their effects will be positive and their risks will be manageable.” Oleg Boguslasvkyi, CTO at Data Science UA, emphasizes that societal attention to AI alignment is rising.

#Defining AI Alignment

At its core, AI alignment addresses ethical considerations to protect humanity from potential existential threats posed by advanced AI. On a less dramatic scale, it encompasses research aimed at ensuring AI systems operate safely and align with human values.

This involves programming AI with complex value systems, fostering fairness, implementing scalable oversight, auditing and interpreting AI models, and preventing harmful behaviors such as power-seeking tendencies.

AI alignment seeks to bridge the gap between the mathematical foundations of large language models (LLMs) and the nuanced social skills humans employ in communication.

As AI technology advances rapidly and AGI becomes a tangible goal, addressing these concerns is crucial. The implications span various domains, from language models to robots and autonomous vehicles, affecting many aspects of daily life.

#AI Alignment in Practice

Generative language models, by default, lack understanding of morals, ethics, human emotions, and the reasons behind ethical constraints, such as not providing instructions for harmful activities.

Fine-tuning processes instill these characteristics in AI models. For instance, ChatGPT, a fine-tuned LLM, can recognize potentially harmful questions and respond appropriately. In 2023, OpenAI committed 20% of its computational resources to AI alignment efforts.

While training materials are carefully selected to exclude explicitly harmful content, the vast corpus used for training may still contain fragments of such information.

#The Two Main Steps of AI Alignment

The first step involves developing specific instructions to ensure LLMs respond correctly. This is achieved by presenting example tasks to the model, which are then refined through human or AI-guided corrections. During this phase, the model also learns to seek clarification if a task is unclear.

The second step begins once the model can adequately address requests. Here, the model generates two responses and solicits feedback from a human or another instructive AI to determine the better answer. The superior response is rewarded and used as a benchmark for future answers.

#Ensuring Truthful Information and Addressing ChatGPT’s “Lies”

The tendency to produce misleading or false answers is a challenge for all LLMs. This issue arises from the vast and sometimes inaccurate information available on the internet, such as content from The Flat Earth Society. LLMs lack critical thinking, making the reward mechanism vital for guiding them towards accurate responses.

Human involvement in the learning process presents another challenge. LLMs generate answers closest to previously rewarded responses. When lacking an answer, an LLM might fabricate plausible information to receive a reward, a phenomenon known as “hallucination.”

Efforts are ongoing to ensure AI systems provide only objective truths and remain honest, crucial steps towards achieving reliable AGI and ASI. Getting the first 80% of accuracy is much easier than achieving other 20%. Nowadays, it looks like hitting 100% accuracy is not going to be possible with LLMs.

#Scalable Surveillance

Current training methods for AI systems incorporate machine ethics, instilling values like equality, impartiality, and non-harm. Despite human monitoring efforts, the process requires modernization to mitigate human error.

To address specific issues, AI-focused companies are incorporating source citation capabilities in LLMs. For instance, OpenAI announced that GPT-4 would include this feature.

OpenAI is also working on developing comprehensive oversight mechanisms to control superhuman AI, aiming to eventually create a superhuman automated AI alignment researcher.

By advancing AI alignment, we strive to ensure that the future of AI remains beneficial and secure for humanity.

CONFERENCE

This year, the GITEX Global 2022 conference was held on 10–14 October. The conference, as always, took place at the Dubai World Trade Center with the participation of TechRadar Pro. Perhaps GITEX is not yet a vastly known conference among Europeans as, say, CES in Las Vegas or Web Summit in Lisbon. Still, the organizers are putting an effort to change that and make GITEX the main technological event of the year not only in the MENA region but also in the whole world.

This was the 42nd GITEX conference, the first exhibition was held in 1981. The popularity and authority of GITEX in the region are so high that the exhibition began to be held in other countries as well: Saudi Arabia, Egypt, Lebanon, and India. In May next year, it will be held in Africa.

GITEX GLOBAL brings together 7 events from technological industries: AI Everything, North Star Dubai, Fintech Surge, Future Blockchain Summit, Marketing Mania, and two new events joining the list in 2022: Global DevSlam and X-VERSE.

I would like to single out AI Everything — one of the largest and most influential events in the world dedicated to artificial intelligence. The AI ​​Everything conference is organized by the UAE Minister of State for Artificial Intelligence, Digital Economy, and Remote Work. Its goal is to promote a more detailed understanding of modern artificial intelligence technologies, strengthen its long-term potential, and expand the global AI ecosystem.

This year, GITEX Global attracted more than 120,000 attendees, 5,000 companies from 170 countries, and more than 1,000 speakers and covered all 186,000 square meters of the Dubai World Trade Center.

The region’s technology leaders have set a course for Dubai’s digital transformation, including improving consumer lifestyles, and government & business services. This year’s conference focused on IoT, digital security, Big Data applications, and cloud technologies. In the field of IoT, local specifics, which are extremely important for the UAE, are fully taken into account.

The smart city is one of the most significant topics of GITEX. Smart cities should be built according to the 4P principle: public, private, people, and partnership. In general, the topics of Smart Things and IoT have given a significant place at the exhibition — both individual suppliers and the largest operators offer their solutions here.

The UAE wants to be not only a financial but also a technological hub in the Greater Middle East region. Bets are placed on technologies such as AR/VR, 3D printing, IoT, smart cities, 5G, and artificial intelligence: these were the most popular new-wave technologies presented at the GITEX booths.

At the same time, the direction on which they are betting the most in the UAE is AI. And Emirates has every chance to take the position of the new technological leader in this breakthrough industry.

The main difference between GITEX and other large technology conferences is that, in addition to giant companies and startups, there is a separate “state” pavilion, where hundreds of state institutions from the entire region, ministers, and public sector officials present the main digital initiatives, innovations, and projects of the year, and also announce public-private technology partnerships and showcase their achievements in the digital transformation of the UAE.

The Dubai Government Pavilion brought together 33 government organizations, including the Ministry of Foreign Affairs, the Dubai Public Prosecutor’s Office, the Department of Tourism and Economy, the Dubai Sports Council, the Department of Education and Human Development, the Department of Legal Affairs, the Department of Finance, the Dubai International Financial Center ( DIFC), Security Industry Regulatory Agency, Dubai Government Human Resources Authority, Dubai Culture and Arts Authority, Community Development Authority, Dubai Health Authority, Aviation Authority, Dubai Airports, Dubai Judicial Institute, and many other government bodies.

Saudi Arabia, Abu Dhabi, Oman, Sharjah, and Bahrain had similar stands.

States and cities are represented at the conference as separate tech companies.

And this breaks the patterns of the bureaucratic system, the black box model of management, which is familiar to most of Europe. Technological progress makes it possible to build a transparent state that is “easy to use” for citizens. And, as far as I’m concerned, it’s a win-win.

Currently, most of the technologies to which the exhibition is dedicated are entering the stage of technological maturity. Therefore, at GITEX, a lot of attention is paid to practical issues of their application — for example, improving the traffic situation after the appearance of smart cars, virtual reality as a means to improve surgery and education, artificial intelligence in various services, including, as already noted, public ones.

THE SCALE OF THE CONFERENCE

If we compare the scale of the conference in numbers, this year GITEX announced itself very loudly and confidently: 120K visitors from 170 countries. For comparison, Web Summit this year will gather 70k, and CES — 80k visitors.

And while the Web Summit (which will take place from November 1 to 4 this year in Lisbon, Portugal) is more tied to technology but less tied to Portugal, GITEX encourages participants to start a business here and work with local companies. CES (which will be held from January 5 to 8 next year in Las Vegas, USA), encourages working specifically with the USA.

The total time of presentations this year was 500 hours. Attendees heard from more than 1,000 government ministers, global technology leaders, and experts who shared their views on key trends in the technology industry.

LIFE HACKS

No matter how much you want to listen to all the speakers, there are only 24 hours in a day. In the midst of the technology conference season, I want to share life hacks on how to get the most out of such events and not get lost in the crowd of people and information.

• My first recommendation is to spend the conference time only on the conference. Reschedule/cancel all non-conference work meetings, calls and questions. Networking and the constant flow of information require a lot of energy and focus.
• Acquaint yourself with the conference before your arrival. Get to know the participants and communicate on your social networks that you will be at the conference, so you can create a list of people you want to meet in order of priority in advance. I usually create a long list and a short list of meetings.
• Think about getting to the venue of the event before it starts. You need to think about this in advance. When 100,000 people go to one place, traffic jams reach their maximum: 10 out of 10. You can spend a lot of time standing in them. It is better to arrive earlier.
• Bring water, charge all your gadgets, and bring your business cards in case your phone goes dead.
• After arriving at the conference, try to go around the location as much as possible on the first day in order to better navigate the place and have an idea of ​​the location.
• The most important thing after the conference is to connect with all the people you talked to. Add to LinkedIn and write a message to everyone, already creating an additional personal connection.
• After busy days filled with constan tmeetings, be sure to give yourself time to recover and plan a rest for at least 1–2 days in silence.
• Attending such conferences can help you build connections with potential partners, and it’s also a great opportunity to learn about new technologies and trends. I wish you great networking and useful acquaintances! Take the maximum and a little more from the conferences.

I hope everyone, who plans to visit Web Summit and CES, will make successful and productive connections!

Artificial intelligence has been continuously developing over the past few decades and has achieved fantastic results and can do unbelievable things. Self-driving cars, face-recognition systems, and virtual assistants are not a fantasy anymore. We can expect even more cool things in the future.

However, as it is widely known, there are multiple myths related to AI, some of which are partly true and even may become real in the future. Others are just beliefs that exist among people today. The latter ones were born as a result of sci-fi movies and books and look more like conspiracy theories.

This article will explore some of the most famous myths, explain the reality behind them, and discuss how AI myths are presented in cinematography.

So, let’s begin!

Myth 1: AI Will Take Our Jobs.

Automation has always been of great interest to people. The first real breakthrough was the industrial revolution of the 18–19th centuries. The use of machines for manufacturing rose dramatically and led to an increase in production. But there was a side effect. Many workers were substituted by machines and lost their jobs.

A similar situation is occurring today where AI is helping to automate many processes. Lots of jobs become obsolete, and people are let go. Several studies were already conducted to determine the positions that are most likely to be automated. Among them: cashiers, drivers, delivery men/women, and even farmers. These are the types of jobs that do not require a whole lot of education or creativity. They are repetitive and constant.

Jobs that require ingenuity and empathy, on the other hand, are much harder to automate. These include teachers, nurses, scientists, architects, engineers, etc. And the good news is that the constant progress of science and technology requires more people like these. Much more workers will be needed to create, build and maintain machines that will do all those routine tasks. These types of jobs are, for sure, more challenging. They call for much attention and education. But they are also more exciting and usually better paid.

The conclusion is that mostly unskilled workers will be substituted for machines. Many jobs may indeed disappear, but new ones will be created while the demand for skilled workers will only increase. This is likely to lead to higher total intelligence and new great inventions.

Myth 2: AI is More Intelligent Than People.

Speaking of intelligence, it is a common misconception that AI is smarter than humans. Many believe that AI’s speed, accuracy, and complexity are proof of its extremely high intelligence. However, the ability to quickly process large amounts of data does not necessarily mean that a machine is intelligent.

A machine cannot do tasks on its own. It must be programmed or trained on a specific task or set of tasks. Therefore, it will only perform well in what it was trained to do. Why is this so?

A computer does not really understand the world in the way humans do. Computers understand numbers. Therefore, everything computers can work with (texts, photos, sounds) is represented by numbers. Also, they do not have any particular meaning — a computer will not understand that it is working with a picture of a dog. It will only see the numbers, make calculations, find patterns, and show the result. The interpretability of the resulting numbers lies on the scientists as well. This is definitely not the kind of intelligence we were talking about.

So far, there were no robots created that can do everything a human can and this is not likely to happen in the near future. As a result, machines are not more intelligent than humans. They are certainly faster. They can process more information. They can find complex hidden patterns in data. But they are still not able to perform well on their own.

Myth 3: AI is Used Only by Large Corporations.

It is a common belief that only huge companies and internet giants use AI. Why would a local shop or a school need it? This notion is fundamentally wrong.

First of all, AI can offer great benefits to any industry: from retail and banking to medicine and the military. Any business, even small local shops, can find something for itself: to increase its sales, improve advertising, diminish customer churn — the list can go on and on. So why not use it?

Secondly, the world is moving towards total automation and digitalization and the role of AI in these processes is exceptionally high. Anybody who misses the chance to adopt it now will incur losses later.

According to this study, two-thirds of organizations have AI programs in place and this number will only increase. As the evolution of artificial intelligence continues, new use cases and further benefits will emerge, and, unfortunately, it may be too late to start using AI then. Therefore the best choice would be to start now and harvest the fruits when others fail.

Myth 4: AI is About Creating Robots.

Artificial intelligence is a broad science that includes many subfields. One of them is indeed robotics, but there are many others. Fields like natural language processing or computer vision are also trendy, and they are rapidly developing.

Many of those who are afraid of AI do not know the reality behind it. Artificial intelligence, in the form of machine learning, is already present in our lives. We may not even know that, but we are interacting with it all the time. Virtual assistants, chatbots, Amazon’s recommendations, and even Google search are all examples of artificial intelligence .

In fact, everything that tries to simulate the way humans think can be called AI. Robotics is an exciting case, but not the only one.

Myth 5: Terminator.

Probably the most widespread myth about artificial intelligence is sometimes called Terminator. Its followers think that AI will rise against us, kill all humans and rule the world. Believers oppose the further development of artificial intelligence and want to stop its existence.

The reality is entirely different. First of all, AI is far from that stage when it will be able to think independently and make deliberate decisions — even the easiest ones. Secondly, there exist so-called laws of robotics that prevent robots from attacking humans. Finally, we should not forget that intelligent and competent scientists are conducting all AI research. They would not be irresponsible enough to allow something like this to happen. Every development is always under strict supervision, and all the necessary precautions are always taken.

As it was mentioned earlier, AI is already among us and, as you can see, there is nothing to worry about. Quite the opposite, we can expect more cool inventions and breakthroughs. The new age is not far away, and it looks beautiful.

AI Myths From Movies

The idea of AI destroying humankind is trendy in cinematography. Many films are based on this theory, starting with Terminator, which actually gave the name to this whole story.

But there are many other movies about artificial intelligence, and not all of them show AI as an evil entity. It is fascinating to look at AI in science fiction, as these fantasies may come true one day. In many ways, fiction is the mirror of our views of the real world, so appearances of alleged AI technologies in the works were destined to become one of the ways we shape our understanding of it. Let us take a look at some of the most exciting stories that the movie industry gave us over the years.

The Matrix: Uprising and Simulating Reality

Arguably the most famous cinematic appearance of AI happened in the classic The Matrix trilogy, which became one of the most popular sci-fi films at the turn of the century.In those movies, AI plays an extremely valuable story point here, raising two big questions simultaneously.

As a trending theme among many fiction works, the conscious machine rebellion is the plot starter in the series. AI has been designed in the twenty-first century, which led to intelligent machines, which obviously came to self-understanding of their status and started an uprising. And humans, with their capabilities, stood no chance against these intelligent machines.

Speaking of intelligent machine technology, researchers around the world are already developing such AI technologies. However, it is essential to understand the difference between real-world intelligent machines and the self-aware units from The Matrix and several other films. Current technology is far from being able to build artificial creations who are able to understand the world around them as closely as humans do. Today machines are rather built to correctly interpret the circumstances of the exact task set to them.

Developing the artificial mind was a high-reach goal for the AI researchers from the very start of the field, but right now, the distance to the possible solution is only if a bit shorter than before.

Ex Machina and Others: Human-like AI

Ex Machina is one of the most famous modern movies about AI that behaves like a human and physically resembles a human. Having featured in lots of films over the years as different kinds of robots or artificial humans, this type of AI usually presents itself as a conscious creation that tries to imitate the human form as closely as possible.

Origins

The entire idea of human-like constructed intelligence has roots much deeper than modern movies. Some of them are directly inspired by earlier literature works. One of the most prominent such examples is I, Robot, released in 2004, which was based on an Isaak Azimov novel from as far back as 1950.

The dreams and ideas of the sentient artificial creation overall are very ancient in world culture and this is the point where we can connect myths about AI with actual Greek myths. In the ancient literary works, artificially crafted people appeared as creations of gods, with a purpose either to serve them or wage battles against their enemies. Such creations often proved to be hostile to humans as the story unfolded, resonating with views from the much newer works.

Human-like AI: A Possibility or a Dream?

But is this possible in reality, or even in the near future? There are two main aspects of this AI concept that are far beyond science at this point.

The first one is machine self-awareness. As pointed earlier, current AI concepts mainly focus on solving the selected tasks, and under the hood, such machines are far from grasping the context surrounding these problems. They mostly are learned enough to match patterns found in the surrounding world with a given set of ideas but are currently unable to understand the deeper context behind these patterns.

Another point that is still far from realization is the physical appearance of artificial creations. Current Artificial Intelligence machines usually employ a virtual form, with different sensors and detectors that take in the world-observing functions. However, in most cases, they are far from imitating what humans feel in similar situations.

Indeed, in recent years some progress has been made in giving different AI systems an overall shape close to feral , as was the case with the infamous Boston Dynamics robots. However, these systems are only replicating the form of the natural behavior or senses, being yet far from grasping the concepts of what lies beneath. Machines are still unable to think as humans, though they already can interpret the world just as profoundly .

In this sense, current robots are much closer in appearance to the TARS robot from Interstellar (2014) than to the human-like visually Ada from Ex Machina. TARS did not look anything like a person while still being a fully communicative and functional human assistant. The technology that would allow us to build robots like TARS is not yet here, and it resembles another way they can be designed: with no unnecessary humane features.

Conclusions

The myths that surround Artificial Intelligence mainly spring from the sharp rise in the field and have endured in the last several years. With fields gaining new applications almost daily, many people still do not fully understand or grasp what current AI is and how it all came to work. And this is where the myths and legends arise and fill the gaps left unanswered by experts in the field.

Either created by the right questions or the popular creations, these ideas often unconsciously formour vision of AI. Therefore, being in touch with current Artificial Intelligence trends is almost a must to understand its development. But don’t forget that sometimes our myths can create dreams, a realization of which can lead to something new and beautiful.