The Sea Wasp (Ryk E. Spoor)

So here's todays discussion. It is in fact related to some personal experiences, but I'm not going into those here. 

( Read more... )



 

Regardless of whether "AI" ever actually becomes intelligent, rather than just being an increasingly good "predict the kind of thing you want or should see out of this query" machine (a completely separate and also increasingly complicated subject), LMM and related "huge trained neural network" AIs are here and people have invested unfathomable amounts of money into them. Even if the bubble explodes, these AIs will still exist. 

So what SHOULD they be used for?

Well, right off we get a conflict between "in the ideal world" and "in the current world". 

You see, an awful lot of the conflict about AI right now -- the copyright suits, the arguments about using it for doing desk work, for finding ways to emulate dead, or even still living but aged, actors, etc. -- REALLY boils down to this:

Our society has no support network. So anything that humans do that sustains them is specifically a matter of SURVIVAL. 

It's not just a matter of writing fun stories or making silly pictures. It's a matter of that being a significant survival element, perhaps the ONLY survival element, that many people have to keep them from disaster. Thus, any device or method that looks to make the individual's contribution to this work less valuable is a direct threat. 

For AI, the problem is that it is QUALITATIVELY, as well as quantitatively, different from prior technological advances. It is GENERALIZABLE to tons of tasks that were until now almost entirely the domain of human endeavor. Trained LLMs are getting better and better at recognizing and copying and adapting multiple different types of writing - not just individual human styles, but different kinds of writing -- professional proposals, book reports, novels, patents, etc. -- and there's a LOT of people that threatens, and the number of people whose jobs are at risk is increasing with every improvement of the technology. 

It is also inherently FAR more deployable for such tasks. If I want to make, say, a Terminator bot, even assuming I have the AI for it available, building a militarily robust, armed, flexible, powerful independent robotic platform is TOUGH, and takes a long time, just like retooling a factory. 

But if your AIs are already generating text and can format it into Word, it takes basically NO effort to replace the guy at the desk with the AI writing software package. 

In the IDEAL world, human survival and basic happy living would be ensured -- the robotic deployment and increase in productivity would be partially diverted to supporting all the people involved. Such people could then write what they wanted, paint what they wanted, with or without AI assistance or interference, and it would not impact their ability to live well. 

That's not the way it currently works, though, so I am very much against the current trend to try to find ways to use AI to displace existing human workers in areas the humans depend on. 

However, there ARE areas in which modern large-trained-neural network systems absolutely can and should be used even now.

For example, AIs are extraordinarily good at pattern discovery, and can also be trained to ANALYZE the patterns to see if a coherent framework emerges. 

This is ideal for things like mathematical and physical/materials research, especially in the theoretical areas or the design realms where much of the problem is that the overall subject area is far, far too huge for a human being to comprehend. An AI properly designed could, at the least, pull out multiple "huh, that's funny" areas in a given field and draw a human's attention to them for further analysis. Some AIs are already showing the ability to perform what appear to be solid mathematical proofs, which is quite an interesting capability and has implications not just for mathematicians but for things like quantum computation and materials design. 

AIs of this nature can also probe and model the structure of an astonishing number of chemical compounds and, perhaps more importantly, metamaterial structures, to discover materials that can do things we didn't know were possible -- or ones we did know were possible, but were having problems finding practical methods to achieve. New antibiotics, perhaps; optical metamaterials with negative indices of refraction; superconductors and super-insulators of both electricity and heat. This is the kind of thing AI is properly made for -- locating patterns within masses of data or of processes that are far too complex for human beings to view as a gestalt. 

The same thing applies to medical advances; understanding the complexities of modern medicine is mindboggling, and what's needed is a way to somehow locate the important anomalies within a vast ocean of data. AI can do that.

Back in the 1700s-1800s, it was possible for one bright person to know pretty much everything in the sciences, and thus be able to make cross-connections between the fields, synthesizing knowledge from the combination. That's an impossible thing for one human being to do now. 

But a human with an AI to help make the connections? That's not ridiculous at all. 



 

Below, with a cut for length...

( Read more... )

Cut for length...

( ... see the actual post! )


 

Specifically to the one here in Troy, NY. 

Troy isn't a large city, so we didn't draw the huge crowds you see in the big metropolises, but there were between 1,000 and 1,500 people crowded into the little Riverfront Park area -- for Troy, that IS huge. 

Attendance was of all ages -- I saw one grandmotherly woman likely ten to twenty years older than me with a sign saying "Why Do I *STILL* Have To Be Marching In Protests?", and there were people the age of my various kids and everything in between. 

There was no violence, no confrontations with ICE or police -- in fact, the police simply watched, kept the roads safe and clear, and made sure everything moved as smoothly as possible, even when the protesters streamed all the way across the Green Island Bridge and back. The organizers took pains to remind the crowd that not only were we nonviolent, our job was de-escalation even if threatened. Fortunately, that caution didn't need to be used. 

The crowd was united in its disapproval of the current regime of lunatics and traitors and showed it in their signs and in their costumes. Scooby-Doo was there, as were the expected T-Rexes, giant frogs, and also a Minion and multiple others. I wore the Straw Hats' Jolly Roger as a cape, plus a straw hat. There were multiple One Piece fans around who commented on it -- one young lady carried a sign that said "The only king I want is LUFFY, The Pirate King!". 

The speakers were for the most part energetic, articulate, and even understandable, which is often a problem for me in public venues. One gentleman dressed in Revolutionary War getup read the various charges against the King, pointing out how Mad King Trump is echoing a lot of Mad King George's offenses. New York Democratic politicians actually showed up, Representative Paul Tonko probably the most notable.

In some ways, it was really like attending a folk festival, including some songs in between speeches. There was anger, but not directed at anyone there, only at the crazy people in Washington who are not only damaging our country, but embarrassing the hell out of all of us. Forget economics, it'll take decades to live down having this bozo in the White House at all. 

Mostly it as filled with positive energy, people here to have their voices heard, and  to be part of a larger movement against our incompetent fascist regime. 

... but that meeting was sure bad. As many have noted, some of the expressions in that room were ones you NEVER want to see on the faces of flag officers. Especially the Commandant of Marines. 

They didn't explicitly go for a loyalty test -- but they did threaten the top eight hundred officers and staff, which is always a great way to endear yourself to the military. 

They didn't present a grand strategy to usher in a new imperialist era... because they're focused on literally sending the army against Democratic cities. 

They didn't have a mass firing... just a lunatic ranting third rate macho bullcrap as new regulation instructions that are specifically targeted in ways that will eliminate a vast number of POC and women from the ranks, while doing nothing at all to actually improve the functioning of the military. 

They didn't announce martial law... but they did announce "Geneva Convention? More like Geneva Suggestion!". 

This was simultaneously frightening and just plain embarrassing. I don't understand how Hegseth, at least, didn't realize how insanely stupid his whole act was. Trump has dementia, so at least he's got an excuse for being unaware of anything around him. 

Gods above and below, what a clown-car of banal horror.  

... they start drilling. 

This time it's the incompetent Secretary of War calling together effectively ALL of the flag officers throughout all branches of the military into a single meeting, in ONE WEEK, publicly announced.

I can't even BEGIN to address all of the reasons this is stupid beyond easy belief, and also highly concerning in other ways, but... holy Jebus. 

First of all, wherever the meeting is held -- barring it being a hidden conference room in Cheyenne Mountain, maybe -- you have just created the biggest damn military strategic and tactical target the modern world has EVER seen. You literally have gathered the ENTIRE chain of command, minus the President (who in this case is worse than useless anyway), into ONE ROOM. A single attack could literally behead the USA's military machine in a few seconds.  

There's a lot of jokes about how the top brass are expendable, and there's always a few that are, but the fact is if you take away HUNDREDS of people at the top of a regimented, strictly-organized military you will create chaos. There's SO many parts of that machine that assume a reasonably smooth flow of information from top to bottom and back up with carefully-designed succession backups. 

But not succession backups fhree or four deep. The PRESIDENT is like a dozen deep, but most officers have one or two people who can step into their places and reasonably well catch the load. A lot of THOSE are people at or near that officer's level. 

If you're taking away ALL the people at the top several levels, there's no backup or precedent for that kind of "beheading" strike. 

So that's ONE level of stupid. 

A level of CONCERNING is that we have no idea of the REASON for this meeting. Why the hell would you suddenly summon every single ranking officer to you? 

I can't think of any reason that's GOOD. This could be a "show us this loyalty" moment, where they basically want to make sure every one of the top brass are behind Trump--- or are removed from office and replaced. 

This could be a global strategy meeting (recognizing that Hegseth's idea of strategy will be barely superior to Trump's) in which they're going to plan some ridiculously overarching plan to get rid of *ALL* of "America's Enemies" in some bargain-basement James Bondian scheme. 

I don't THINK it could be a Vlad Tepesch situation -- we're not quite back in the quaint era of executing all your enemies in a single ballroom -- but I am absolute stymied trying to imaging a halfway sane reason for doing this. 


Our current government, ladies, gentlemen, friends, foes: a government of the grifted, by the grifters, for the grifters, run by clowns that would embarrass Pennywise and the Joker. 

There was an ongoing problem with Amazon's listings of my space opera, Demons of the Past (Revelation, Revolution, Retribution), making the three of them have the same cover and confusing would-be purchasers. 

Finally it's straightened out, and here's the proper links:

#1, Revelation: https://www.amazon.com/Demons-Past-Revelation-Ryk-Spoor-ebook/dp/B0DN6Q6ZZ7
#2, Revolution: https://www.amazon.com/Demons-Past-Revolution-Ryk-Spoor-ebook/dp/B0DN6SVXF6
#3, Retribution: https://www.amazon.com/Demons-Past-Retribution-Ryk-Spoor-ebook/dp/B0DN6PL1ZT

This is the work I spent the most hours on over the years, rewriting it at least five times since its first draft in about 1982 or so (and the idea having been first thought of in the late 70s).

I've been notified by three people (so far) that I'm now _persona non grata_ because of a post I made regarding the killing of Mr. Charlie Kirk yesterday. 

Two of them made statements that clearly imputed to me statements that I hadn't made, but that they had apparently inferred from what I HAD said. 

This is unfortunate, and I don't expect to change their minds (or, in general, anyone's mind, online) about such things. They've made their judgment, they now have that perception of me, and arguing about how someone perceives you is usually a lost cause from the start.


But for the record, I try to write *exactly* what I mean. If I mean to say "Ho, he deserved to die, good job!", I'd quite literally post exactly that. 

What I posted said that I don't approve of killing people as a solution to the problem and I'd like to live in a world where that's not viewed as an appropriate solution. 

I did then note the irony of the fact that he had explicitly said (very shortly after a school shooting) that some gun deaths were a necessary price for maintaining the Second Amendment's protections. 

I also noted that I wouldn't waste prayers, if I prayed, on him, given what he promoted in life.


NONE of that says "I approve of killing people who disagree with me politically". If I want to say that, I don't need to type that long (especially on FB from a phone, which is a big PITA). I can say "Shooting people like Charlie Kirk is a public service and we need more public servants". 

I don't try to hide my beliefs in my fiction OR my nonfiction. The closest I get to "subtle" (aside from hiding little Easter Eggs in the Arenaverse) is when I had Jason Wood make an anti-Patriot Act speech thinly disguised as a protest against werewolf-triggered paranoia (since the Morgantown Event is basically his world's 9/11). 

If I want to say something, I say it. And I say it very carefully. 

If I DON'T say a particular thing, odds are excessively strong that I don't, in fact, mean that thing. 

Again for the record, no, I don't approve of people shooting people under any circumstances aside from actual self-defense (he's coming at me with intent to injure or kill). I don't even approve of it in wartime, though by the nature of the beast it does and will happen and I'm generally not going to judge the soldiers for it. 

I think Charlie Kirk was doing the world a lot of disservice, and I wish he hadn't done some of the things he was doing, but that didn't earn him a bullet nor do his family and friends deserve the shock. 

At the same time, he as an individual leaves me with no particular fond feelings and I feel no obligation to pretend about it. 

And I find it grimly, ironically amusing that he publicly espoused the "necessity" of some number of gun deaths to protect the Second Amendment.

This is not, in any way, an approval of the killer or of assassination in general. 

I am UNSURPRISED that such things are happening -- to people on both sides of the aisle. 

This event and some discussion after it, though, did give me a different epiphany, which I'll write about separately.  

The SFWA (Science Fiction Writers of America) has announced that they are participating in a class-action lawsuit against Anthropic AI, which used an absolute metric shitton of authors' books to train its AI. While it's been ruled in one case that these actions don't, technically, constitute copying the book (because the training doesn't leave actual copies of the trained books, only of the  responses to having been trained on it, in short), it HAS been ruled that just grabbing copyrighted material and using it for a commercial purpose (such as training your commercial AI) is not a fair use. 

Anthropic AI is currently valued at around 150-160 billion dollars, just as a note. This is not a small company. 

From my point of view, it's absolutely open and shut: did they make use of copyrighted works to make a commercial product? Yes. Did they know they were doing so? Yes. Did they know they SHOULD pay for the rights to make use of those works? Yes. They simply concluded that it would be expensive, so they grabbed archives of pirated copies. 

The penalties for this should be substantial. This isn't like someone just downloading a book to read, in which case the most you could argue is that they owe you the purchase price for the copy they made. This is taking people's copyrighted work to use to make a commercial product that you intend to profit from. Conceptually this is no different than making a movie or other derivative work from the copyrighted material. The movie may differ drastically from the book -- it may in the worst case have little but names to show the connection. Even so, the moviemaker HAS to have paid the author for the rights to make the movie using their book. 

Note that there is no argument in this case that Anthropic did not, in fact, make use of these works. It's admitted that they did. 

But if "not retaining a copy, just the impressions" is good enough, then why can't I go and publish a Lord of the Rings fanfic? If I put the book away and don't look at it while writing, I'm just using my own impressions from the book to write the fanfic. Better yet, there's a lot of books I've only read once; if Anthropic's allowed this argument, then I should be able to freely use anything I remember from any book I've ever read. 

To an extent, of course, we DO do that -- we're influenced by everything we read, inspired or angered by it. But we also are expected to make a conscious effort to not merely TAKE the intellectual property. Since current AIs are incapable of "conscious effort", and by their nature literally do not RECALL the sources of their training (part of Anthropic and others' defense against accusations of 'copying'), the responsibility for such conscious effort devolves upon Anthropic and their personnel. 

Thus, it would be my contention that Anthropic currently owes every author whose work was used for this training, first a licensing fee -- negotiated appropriately for current and anticipated valuation of their business -- and second, a penalty fee for having DELIBERATELY chosen to try to avoid doing the legally obvious and required licensing. 

I would think that a minimum for that would be a thousand dollars per book infringed for licensing, and five hundred for being deliberately sneaky about it. That's a lowball figure -- note that even an OPTION to use someone's book for a movie -- not even an actual rights assignment -- is usually in the thousand-plus range. In this case it's not just an option -- they DID use the intellectual property. 

The other reason it has to be a significant number is that everyone is aware that the various IP industries are very much interested in eventually using AI to supplement or even replace human creators. If that's the goal, well, those of us who'll be being used to TEACH our replacements deserve a hell of a salary, so to speak. 

I hope this suit goes forward well.