And here I was already to eek out an existence fixing things inside a giant AI complex for food pelts, like some sorta gut bacteria.
Meh guess I will just go back to quitely waiting for either global warming, or some easily treatable disease that my insurance won’t cover, or some religious nutbag to shot me.
We’re just seeing photoreal video, barely two years after blurry images with too many fingers. At a bare minimum we’re going to get an explosion of effortless CGI. At the extremes we might watch the entire film industry collapse.
Being able to coax a finished shot out of your computer, just by describing it into existence, will not coexist with streaming services that deny writers a third season. People will tell stories. It will look like they have actors and sets and cinematographers. They will not. (Though they’ll probably still need editors, because you can’t save bad writing, and a bad edit can ruin anything.)
And that’s the boring use of these tools. Feeding them art halfway through the process will let any one person crank out a whole-ass cartoon by drawing a couple frames. You can still have actors and sets and cinematographers, and then let the computer fill in a magical city outside the window, and make that alien costume’s third arm believable. Text descriptions are the least you can put in - not the most.
The applications in things that matter are similarly wild. We’ve basically invented computers that do what they’re told. More Douglas Adams than Isaac Asimov at this point, and not embodied enough to solve or cause real-world problems on their own, but systems where people can forget they’re not dealing with a human. Calling this brief early period of weird shit starting a “peak” is completely ridiculous, even if all you’re going by is the Rich People Feelings Graph.
There are a few reasons why the AI hype has diminished. One reason is data integrity concerns - many companies prohibit the use of ChatGPT out of fear of OpenAI training their models on confidential data.
To combat this the option is to provide LLMs that can be run “on premise”. Currently those LLMs aren’t good enough for most uses. Hopefully we will get there in time, but at this pace it seems like it’s taking longer than expected.
Author doesn't seem to understand that executives everywhere are full of bullshit and marketing and journalism everywhere is perversely incentivized to inflate claims.
But that doesn't mean the technology behind that executive, marketing, and journalism isn't game changing.
Full disclosure, I'm both well informed and undoubtedly biased as someone in the industry, but I'll share my perspective. Also, I'll use AI here the way the author does, to represent the cutting edge of Machine Learning, Generative Self-Reenforcement Learning Algorithms, and Large Language Models. Yes, AI is a marketing catch-all. But most people better understand what "AI" means, so I'll use it.
AI is capable of revolutionizing important niches in nearly every industry. This isn't really in question. There have been dozens of scientific papers and case studies proving this in healthcare, fraud prevention, physics, mathematics, and many many more.
The problem right now is one of transparency, maturity, and economics.
The biggest companies are either notoriously tight-lipped about anything they think might give them a market advantage, or notoriously slow to adopt new technologies. We know AI has been deeply integrated in the Google Search stack and in other core lines of business, for example. But with pressure to resell this AI investment to their customers via the Gemini offering, we're very unlikely to see them publicly examine ROI anytime soon. The same story is playing out at nearly every company with the technical chops and cash to invest.
As far as maturity, AI is growing by astronomical leaps each year, as mathematicians and computer scientists discover better ways to do even the simplest steps in an AI. Hell, the groundbreaking papers that are literally the cornerstone of every single commercial AI right now are "Attention is All You Need" (2017) and
"Retrieval-Augmented Generation for Knowledge -Intensive NLP Tasks" (2020). Moving from a scientific paper to production generally takes more than a decade in most industries. The fact that we're publishing new techniques today and pushing to prod a scant few months later should give you an idea of the breakneck speed the industry is going at right now.
And finally, economically, building, training, and running a new AI oriented towards either specific or general tasks is horrendously expensive. One of the biggest breakthroughs we've had with AI is realizing the accuracy plateau we hit in the early 2000s was largely limited by data scale and quality. Fixing these issues at a scale large enough to make a useful model uses insane amounts of hardware and energy, and if you find a better way to do things next week, you have to start all over. Further, you need specialized programmers, mathematicians, and operations folks to build and run the code.
Long story short, start-ups are struggling to come to market with AI outside of basic applications, and of course cut-throat silicon valley does it's thing and most of these companies are either priced out, acquired, or otherwise forced out of business before bringing something to the general market.
Call the tech industry out for the slime is generally is, but the AI technology itself is extremely promising.
What we have done is invented massive, automatic, no holds barred pattern recognition machines. LLMs use detected patterns in text to respond to questions. Image recognition is pattern recognition, with some of those patterns named things (like “cat”, or “book”). Image generation is a little different, but basically just flips the image recognition on its head, and edits images to look more like the patterns that it was taught to recognize.
This can all do some cool stuff. There are some very helpful outcomes. It’s also (automatically, ruthlessly, and unknowingly) internalizing biases, preferences, attitudes and behaviors from the billion plus humans on the internet, and perpetuating them in all sorts of ways, some of which we don’t even know to look for.
This makes its potential applications in medicine rather terrifying. Do thousands of doctors all think women are lying about their symptoms? Well, now your AI does too. Do thousands of doctors suggest more expensive treatments for some groups, and less expensive for others? AI can find that pattern.
This is also true in law (I know there’s supposed to be no systemic bias in our court systems, but AI can find those patterns, too), engineering (any guesses how human engineers change their safety practices based on the area a bridge or dam will be installed in? AI will find out for us), etc, etc.
The thing that makes AI bad for some use cases is that it never knows which patterns it is supposed to find, and which ones it isn’t supposed to find. Until we have better tools to tell it not to notice some of these things, and to scrub away a lot of the randomness that’s left behind inside popular models, there’s severe constraints on what it should be doing.
engineering (any guesses how human engineers change their safety practices based on the area a bridge or dam will be installed in? AI will find out for us), etc, etc.
I haven’t seen that. Sent the same pieces of infrastructure equipment to pure 3rd world to super wealthy cities. Not saying it doesn’t exist but I personally haven’t seen it. You know a lot of these specs are just recycled. To the extent that I have even seen the same page references across countries. Was looking at a spec a few days ago for Mass. that was word for word identical to one I had seen in a city in British Columbia.
In terms of biases among P.E.s what I have seen is a preference for not doing engineering and instead getting more hours. Take the same design you inherited in the 80s, make cosmetic changes, generate more paperwork (oh this one part now can’t be TBD it has to be this specific brand and I need catalog cuts/datasheet/certs/3 suppliers/batch and lot numbers and serial numbers/hardcopy of the CAD of it…). So I imagine a LLM trained on these guys (yes they are always guys) would know how to make project submittals/deliverables longer and more complex while feeling the urge to conduct more stakeholder meetings via RFIs.
Sorry don’t mean to be bitter. I have one now demanding I replicate an exact copy of a control system from the early 80s with the same parts and they did not like when I told them that the parts are only available on eBay.
Please god I hope so. I don’t see a path to anything significantly more powerful than current models in this paradigm. ANNs like these have existed forever and have always behaved the way current LLMs do, they just recently were able to make them run somewhat more efficiently with bigger context windows and training sets which birthed GPT3 which was further minimally tweaked into 3.5 and 4 among others. This feels a whole lot like a local maxima where anything better will have to go back down through another several development cycles before it surpasses the current gen.
I think GPT5 will be eye opening. If it is another big leap ahead then we are not in this local maxima, if it is a minor improvement then we may be in a local maxima.
Likely then the focus will shift to reducing hardward requirements for inforarance(?) Allowing bigger models to run better on smaller hardware
It takes what, a year minimum to design a chip? I think the iterative hardware-software cycle is just now properly getting a foothold on the architecture. The next few years are going to, at minimum, explore what lavishly-funded, purpose-built hardware can do for the field.
It’ll be years before we reach any kind of maximum. Even if the software doesn’t improve at all, which is unlikely, better utilization alone will make significant improvements on performance.
It feels a whole lot more like there are big things due with GPT5 and beyond.
Like, to be a successful actor in 2020 was to act, and stand in front of expensive equipment operated by specialized operators, with directors, makeup, catering… The production itself was/is its own production.
I predict that to be a successful actor in 2030, all you’ll need is a small amount of money to utilize some powerful processors over the internet, enter in a few photos of your face, give it 10 different ideas for a movie, until it to make some 2 hour films where you are the star. Then you’ll take one of them that you kind of like, throw some prompts at it and end up with a nearly finished Hollywood quality film.
To be a successful musician in 1960, you needed to get a record deal, you needed to go to a recording studio. Now we’ve got Jacob Collier winning Grammys, recording everything in his bedroom. I think we’re going to see that kind of history repeat itself on steroids. Not just for art, though. For anything.
With the rapid advancements were seeing in robotics right now, I can’t imagine a single thing that people do that won’t be done better by autonomous agents, both programmatic and robotic, in the next 5 years or so.
I predict that to be a successful actor in 2030, all you’ll need is a small amount of money to utilize some powerful processors over the internet, enter in a few photos of your face, give it 10 different ideas for a movie, until it to make some 2 hour films where you are the star. Then you’ll take one of them that you kind of like, throw some prompts at it and end up with a nearly finished Hollywood quality film.
It's some weird semantic nitpickery that suddenly became popular for reasons that baffle me. "AI" has been used in videogames for decades and nobody has come out of the woodwork to "um, actually" it until now. I get that people are frightened of AI and would like to minimize it but this is a strange way to do it.
At least "stochastic parrot" sounded kind of amusing.
Yeah people have absolutely been contesting the use of the term AI in videogames since it started being used in that context, because it’s not AI.
It didn’t cause the stir it does today because it was so commonly understood as a misnomer. It’s like when someone says they’re going to nuke a plate of food - obviously nuke in this context means something much, much, much less than an actual nuke, but we use it anyway despite being technically incorrect cuz there’s a common understanding of what we actually mean.
Marketing now-a-days is pitching LLMs (microwaves) as actual AI (nukes), but the difference is people aren’t just using it as intentional hyperbole - they think we have real, actual AI.
If/when we ever create real AI, it’s going to be a confusing day for humanity lol “…but we’ve had this for years…?”
Not sure if you're aware of this, but stuff like that has already happened, (AIs questioning their own existence or arguing with a user and stuff like that) and AI companies and handlers have had to filter that out or bias it so it doesn't start talking like that. Not that it proves anything, just bringing it up.
Well, do we do that? Unlike software we can make a much better argument that we deserve rights and should not be slaves. Nothing is really stopping, besides the end of the universe, a given piece of code from “living” forever it shouldn’t matter to it if it spends a few million years helping humans cheat on assignments for school. However for us we have a very finite lifespan so every day we lose we never get back.
So even if for some weird reason people made an AGI and gave it desires to be independent it could easily reason out that there was no hurry. Plus you know they don’t exactly feel pain.
Now if you excuse me I have to go to bed now because I have to drive into work and arrive by a certain time.
Um, actually clueless people have made “that’s not real AI” and “but computers will never …” complaints about AI as long as it has existed as a computing science topic. (50 years?)
Chatbots and image generators being in the headlines has made a new loud wave of complainers, but they’ve always been around.
It's exactly that "new loud wave of complainers" I'm talking about.
I've been in computing and specifically game programming for a long time now, almost two decades, and I can't recall ever having someone barge in on a discussion of game AI with "that's not actually AI because it's not as smart as a human!" If someone privately thought that they at least had the sense not to disrupt a conversation with an irrelevant semantic nitpick that wasn't going to contribute anything.
LLMs as AI is just a marketing term. there’s nothing “intelligent” about “AI”
Yes there is. You just mean it doesn’t have “high” intelligence. Or maybe you mean to say that there’s nothing sentient or sapient about LLMs.
Some aspects of intelligence are:
Planning
Creativity
Use of tools
Problem solving
Pattern recognition
Analysts
LLMs definitely hit basically all of these points.
Most people have been told that LLMs “simply” provide a result by predicting the next word that’s most likely to come next, but this is a completely reductionist explaining and isn’t the whole picture.
Edit: yes I did leave out things like “understanding”, "abstract thinking ", and “innovation”.
Other than maybe pattern recognition, they literally have no mechanism to do any of those things. People say that it recursively spits out the next word, because that is literally how it works on a coding level. It’s called an LLM for a reason.
they literally have no mechanism to do any of those things.
What mechanism does it have for pattern recognition?
that is literally how it works on a coding level.
Neural networks aren’t “coded”.
It’s called an LLM for a reason.
That doesn’t mean what you think it does. Another word for language is communication. So you could just as easily call it a Large Communication Model.
Neural networks have hundreds of thousands (at the minimum) of interconnected layers neurons. Llama-2 has 76 billion parameters. The newly released Grok has over 300 billion. And though we don’t have official numbers, ChatGPT 4 is said to be close to a trillion.
The interesting thing is that when you have neural networks of such a size and you feed large amounts of data into it, emergent properties start to show up. More than just “predicting the next word”, it starts to develop a relational understanding of certain words that you wouldn’t expect. It’s been shown that LLMs understand things like Miami and Houston are closer together than New York and Paris.
Those kinds of things aren’t programmed, they are emergent from the dataset.
As for things like creativity, they are absolutely creative. I have asked seemingly impossible questions (like a Harlequin story about the Terminator and Rambo) and the stuff it came up with was actually astounding.
They regularly use tools. Lang Chain is a thing. There’s a new LLM called Devin that can program, look up docs online, and use a command line terminal. That’s using a tool.
That also ties in with problem solving. Problem solving is actually one of the benchmarks that researchers use to evaluate LLMs. So they do problem solving.
To problem solve requires the ability to do analysis. So that check mark is ticked off too.
Just about anything that’s a neutral network can be called an AI, because the total is usually greater than the sum of its parts.
Edit: I wrote interconnected layers when I meant neurons
The term "artificial intelligence" was established in 1956 and applies to a broad range of algorithms. You may be thinking of Artificial General Intelligence, AGI, which is the more specific "thinks like we do" sort that you see in science fiction a lot. Nobody is marketing LLMs as AGI.
yeah, I guess thats how I was interpreting it. dunno, I see a lot of articles about how its super easy to crack these LLMs using outside of the box thinking (ascii art text to get instructions on how to make a bomb, etc). that doesnt really scream “intelligent” to me.
You imply that humans cannot be tricked by out of the box thinking? Any hacker would tell you that the most reliable method of entry into any system is just ActLikeYouBelong.
Add comment