Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

The Intelligence Age

ia.samaltman.com

Post reply on HN
Print view View original

321–330 of 447 posts

Re: The Intelligence Age

#321

Post by stonogo »

saurabhchalke wrote: post #318
gwern wrote: post #150

> If we want to put AI into the hands of as many people as possible, we need to drive down the cost of compute and make it abundant (which requires lots of energy and chips). If we don’t build enough infrastructure, AI will be a very limited resource that wars get fought over and that becomes mostly a tool for rich people. Translation for the rest of us: "we need to fully privatize the OA subsidiary and turn it into…

Why can't we just go with Occam's razor, and assume that they really believe in their mission of providing access to frontier AI as freely to the world as possible?

Because that doesn't account for the non-profit rugpull he orchestrated earlier in the company's history.

Re: The Intelligence Age

#322

Post by Terr_ »

Der_Einzige wrote: post #276
swyx wrote: post #31

Earlier quoted context omitted.

you need some amount of irrational definite optimism + knowing things others dont to be a good founder. that kind of reality distortion field is why sam is sam and we are here debating phrases on an orange website.

Related, I tongue-in-cheek believe that something analogous to the actual SCP object for a "reality distortion field" may in fact exist. There is zero good explanation for "Teflon Don" or the North Carolina Lieutenant Governor getting away with all the stuff they do while Al Franken got politically crucified.

The least-magical answer for that is that some people have fundamentally different ways of approaching the world, and certain things will be tolerated by certain sets of supporters.

Re: The Intelligence Age

#323

Post by roenxi »

gwern wrote: post #150

> If we want to put AI into the hands of as many people as possible, we need to drive down the cost of compute and make it abundant (which requires lots of energy and chips). If we don’t build enough infrastructure, AI will be a very limited resource that wars get fought over and that becomes mostly a tool for rich people. Translation for the rest of us: "we need to fully privatize the OA subsidiary and turn it into…

And the reason we can be certain that it is an accurate one is that the only way to put AI in the hands of as many people as possible is to commercialise the tech. Non-profits are good at many things, but not for spreading technical improvements.

There aren't many alternatives here. It is commercialisation or looming irrelevance.

Re: The Intelligence Age

#324

Post by jerrygenser »

drooby wrote: post #299

This morning I was reviewing some code that a JR engineer submitted. It had this wild logical conditional with twists and turns, negations, and weird property naming.. o1-preview perfectly evaluated the conditional and determined that, hilariously, it would always evaluate to true. o1 untangled the spaghetti, and, verifying that it was correct was quick and easy. It created a perfect truth table for me to visualize.…

It's surprising that the JR engineer who presumably has access to chat jippity submitted this bad code to begin with. Wouldn't they have had the AI review the code first?

Re: The Intelligence Age

#325

Post by Terr_ »

throw310822 wrote: post #303
moffkalast wrote: post #226

Earlier quoted context omitted.

As large as the absolute largest models are today, they are still microscopic compared to our brains. A 1.7T param model would only store an actual total of about 850 GB if fully saturated (4 bits of information per weight estimated for bf16 transformers), a lot less than a human brain with 150T synapses running in full analog precision. We need to scale the current gen of models at least another 10-100x to even reac…

> 150T synapses running in full analog precision Analog systems are not known for being very precise- they're noisy, signals get corrupted easily- and that's why we prefer digital ones. As soon as we had the technology, we switched everything we could- audio and video recording, telephone calls, photography, to a digital medium. This makes me wonder if the seemingly extraordinary efficiency of artificial neural netwo…

With respect to brain activity, how do we know its really noise, and not just layers of meaning--or at least purpose--which we don't yet understand?

If straightforward binary signaling was so universally superior, I think the worldwide network of over a quintillion ruthlessly self-replicating nanobots would be using a much more heavily after the last billion years.

Re: The Intelligence Age

#326

Post by empath75 »

KETHERCORTEX wrote: post #289
peppermill wrote: post #173

Earlier quoted context omitted.

It's amazing how he can imagine wars being fought over AI but not wars being fought over resources needed to "build enough infrastructure."

> he can imagine wars being fought over AI I wonder how did he come to such a prediction. When was the last time we had a war over advanced tech? Armies didn't fight over telegraph, radio, phones and cars. A war to get AI would also be foolish. A few hundred bombs from your adversaries and AI won't have electricity to function.

It's arguable that every major conflict of the 20th century was over resources required by the combustion engine -- fossil fuels and rubber, in particular.

Re: The Intelligence Age

#327

Post by TOMDM »

spencerchubb wrote: post #294
d_burfoot wrote: post #243

OAI's achievements are amazing. But here's a bit of a skeptical take: cheap human-style intelligence won't have a huge impact because such intelligence isn't really a bottleneck today. It's a cliche that the brightest minds of the age are dedicated to selling more ads or shuffling stock ownership around at high velocity. Anyone who's worked at a big tech company knows the enormous ratio of talent to real engineering…

To raise $1e7, you have to be able to convince VCs that you can make at least $1e9. They want 100x returns from their winners, and don't care about anything else in their portfolio. If instead you only need $1e3 to build it, it doesn't have to make as much money. I could just fund it out of pocket

That and you can justify sinking 100PhD years into a 1/1,000,000

> cheap human-style intelligence won't have a huge impact because such intelligence isn't really a bottleneck today.

What a failure of imagination.

Re: The Intelligence Age

#328

Post by jampekka »

saurabhchalke wrote: post #318
gwern wrote: post #150

> If we want to put AI into the hands of as many people as possible, we need to drive down the cost of compute and make it abundant (which requires lots of energy and chips). If we don’t build enough infrastructure, AI will be a very limited resource that wars get fought over and that becomes mostly a tool for rich people. Translation for the rest of us: "we need to fully privatize the OA subsidiary and turn it into…

Why can't we just go with Occam's razor, and assume that they really believe in their mission of providing access to frontier AI as freely to the world as possible?

Occam's razor says they are saying whatever gives them as much money and power as possible.

OpenAI is suspiciously bad in providing access to their models, compared to e.g. Meta, Mistral, BLOOM collaboration, Alibaba and even oil tyrants FFS.

Re: The Intelligence Age

#329

Post by skydhash »

famouswaffles wrote: post #172
DanHulton wrote: post #131

Earlier quoted context omitted.

The emperor truly has no clothes here. Six months ago, he probably could have gotten away with saying this and there would have likely have been enough people who were still impressed enough with the trajectory of LLMs to back him on it. But these days, most of us have encountered the all-too-common failure mode where the LLM shows its hand, that it doesn't truly understand anything, and that it's just _very very goo…

1. What failure mode do LLMs have that proves they don't understand anything at all ? And why can't i prove the same with humans (who have an abundance of failure modes) 2. You genuinely think that a system whose goal is to predict the data it's given and continues to improve is limited in what it can learn ? Of all the shortcomings of the Transformer architecture, its objective function is not one of them.

> What failure mode do LLMs have that proves they don't understand anything at all ?

Try to get it to write something in a programming language not commonly used on the internet, say Forth or Brainfuck, with only the specifications of said languages. Humans are able to grasp the law of reality through a model and use it to act upon the real world.

> You genuinely think that a system whose goal is to predict the data it's given and continues to improve is limited in what it can learn?

Not GP, but Image generators have ingested more images that I've seen in my life and still can't grasp basic things like perspective or anatomy. Things that people can learn from a book or two. And there are software that already have models for both.

Re: The Intelligence Age

#330

Post by tikhonj »

jagrsw wrote: post #197
AnimalMuppet wrote: post #34

Earlier quoted context omitted.

Even "we will solve the remaining problems" is... perhaps unduly optimistic. At a minimum, we could ask for the evidence.

I'm not here to defend sama, but certain things cannot be proven until they arrive - they can only be extrapolated from existing observations and theoretical limits. Imagine the Uranium Committee of early 40's, where Szilard and others were babbling about 10kg of some magical metal exploding briefly with the power of a sun, with the best evidence being some funky trail in an alcohol vapor chamber. Maybe sama is right…

Hah, atomic power is a great point of comparison: people in the "atomic age" expected atomic power to be everywhere. Electricity too cheap to measure, cars/planes/appliances all powered by small nuclear reactors... That's without going into the real nonsense like radium toothpaste.

And here we are today where nuclear energy is limited to nuclear weapons, a small set of military vehicles and <10% of the world's electricity production. Not nothing, sure, but nothing like past predictions either.

Post reply on HN