The sigmoids won't save you

https://www.astralcodexten.com/p/the-sigmoids-wont-save-you

by Tomte • 2 months ago

308 286 2 months ago

But they do explain the improvement of AI driving 2017-2021 vs 2022-2026.

A lot of words to say "The initial part of a sigmoidal curve is not very informative about the parameters of the sigmoid function in question."

inglor_cz5• 2 months ago

That is true, but I generally enjoy reading a lot of words from Scott, who has a talent for writing.

The entire plot of the Lord of the Rings could probably be compressed into less than 10 kB of text too.

Edit: this seems to be a controversial comment, but IMHO a blog of Scott Alexander's type is an art form, not just a communication channel.

jeffreyrogers• 2 months ago

I find him more interesting when he talks about non-AI topics. Lots of other interesting people are like this too. I'd rather get my knowledge on AI from people who have unique insights into it. Scott has a lot of unique perspectives of his own, but his views on AI are bog-standard for his social group.

andai• 2 months ago

Well, curve shape aside, the high watermark might be lower than where it tapers off.

https://news.ycombinator.com/item?id=46199723

BoredPositron2• 2 months ago

If you use the log scale you'll see that the time horizon of opus 4.6 was as expected...

afthonos2• 2 months ago

As expected by the exponential. The Wharton study was predicting when the exponential would turn into a sigmoid.

ReptileMan• 2 months ago

Everything is linear on a log log scale with a fat marker.

gm678• 2 months ago

I don't know what the Y-axis is supposed to be on that Wharton AI capabilities graph, but I am not really convinced that Opus 4.6 has more than double the intelligence/capability/whatever of GPT 5.1 Max.

BoredPositron2• 2 months ago

https://metr.org/time-horizons/ on linear scale. Clickbait garbage article as most of his in the last year.

afthonos2• 2 months ago

…yeah, that’s where you see the exponential?

NitpickLawyer• 2 months ago

IIRC that graph tracks capabilities as time_to_solve a task for humans (i.e. the model can now handle tasks that usually take a human ~8h). Which, depending on what tasks you look at, could be a reasonable finding. I could see Opus 4.6 handling tasks that take ~8h for humans, and that 5.1 couldn't previously handle (with 5.1 being "limited" at 4h tasks let's say). It is a bit arbitrary, but I think this is what they're tracking.

lukan2• 2 months ago

"It is a bit arbitrary, but I think this is what they're tracking."

I don't know if they can get their numbers right this way, but this seems a way more useful metric, than theoretic capabilities.

jrumbut• 2 months ago

Without knowing more about their methodology, it seems like a lot of the recent improvements have involved the AI itself taking time to complete the task.

At first the models turned a 5 minute task into a 5 second task (by 5 seconds I mean a very short amount of time, not precisely 5 seconds). Then they turned a 15 minute task into a 5 second task.

Opus 4.6 completes 8 hour tasks all the time but (at least in my experience) it isn't spitting the answer out in 5 seconds anymore. It's using chain of thought and tools and the time to completion is measured in minutes or maybe hours.

In my experiments with local LLMs, a substantial part of the gap between frontier and local (for everyday use) is in tooling and infrastructure.

That is why I am sympathetic to the idea we are leveling off. But to bring in the air speed example from the article, I don't think we've reached the equivalent of the ramjet yet. I suspect in the coming years there will be new architectures, new hardware, and new ways to get even more capable models.

MadxX793• 2 months ago

I don't know why people are so impressed by 8h.

I trained an LLM to write the whole Harry Potter series, and that took JK Rowling like 17 years.

For my next point on the graph, I'll train the LLM to write the Bible, something that took humans >1500 years.

myhf• 2 months ago

According to this article: whenever someone games a benchmark to make an upward chart on some y-axis, it's YOUR responsibility to prove how and why that trend can't continue indefinitely.

emoji face with eyes rolling upward

AnimalMuppet• 2 months ago

I'm pretty sure that gaming benchmarks can continue indefinitely.

skybrian3• 2 months ago

Seems to me that the default is "I don't know what's going to happen" and if you're making a confident prediction, bring evidence.

Scott makes a Lindy effect argument which is plausible, but don't let that fool you, we still don't know what's going to happen.

strken• 2 months ago

Check out Re-Bench and HCAST.

The tasks are obviously all of the form "Go do this, and if you get the following output you passed". Setting up a web server apparently takes 15 minutes for a human, which is news to me since I'm able to search for https://gist.github.com/willurd/5720255, find the python one-liner, and copy it within about ten seconds.

Anyway, this is cool but it does not mean Claude can perform any human tasks that take less than 8 hours and are within its physical capabilities.

throwaway274484• 2 months ago

more than double the intelligence/capability/whatever

I'm curious what people really mean when they say this. Intelligence is famously hard to define, let alone measure; it certainly doesn't scale linearly; it only loosely correlates to real-world qualities that are easy to measure; etc. Are you referring to coding ability or...?

adw• 2 months ago

https://podcasts.apple.com/us/podcast/machine-learning-stree... is a pretty good primer on METR, what it measures, and its limitations.

devmor2• 2 months ago

"Exponentials all tend to become sigmoids but you can't predict exactly when" is a true statement, but I'm not sure it needed an article.

This doesn't say much, and the author fights their own points a couple times, suggesting that they maybe didn't think through what they wanted to write until they were in the middle of writing it and started realizing their assumptions didn't match what they expected the data to say.

I really don't get the point of what I just read.

aspenmartin5• 2 months ago

The point is the tiring arguments from AI skeptics saying “things are flattening, they have to” which while technically correct says nothing because no one knows when that will happen and we see no mechanism for this yet. Lindy’s law as a reasonable prediction under total uncertainty is interesting and insightful and a lot of people don’t know about it or why it holds. I did enjoy the reference to this!

devmor2• 2 months ago

But those skeptics are initially responding to the constant AI hype claims that we are exponentially growing to AGI. So this article is in fact just a (very poorly thought through) attempt at saying “nuh uh, the hype might be true, you can’t prove it’s not yet!

solid_fuel3• 2 months ago

Nah this is making a category error. You're assuming that AI skeptics agree that models are demonstrating intelligence along the same axis as humans and that with further improvement they will become equivalent to humans. I am an AI skeptic, and I disagree with this assessment.

Model reasoning is on an s-curve, which is improving.

Model intelligence is not the same as reasoning. It's a different axis, and one I have not seen much movement on.

See, humans have a recursive form of intelligence which is capable of self-reflection and introspection. LLMs can only reason about tokens which have already been emitted. Humans and LLMs do not share the same form of reasoning, and general human-like intelligence will not arise from the current architecture of LLMs. Therefore it is a mistake to assume that continual improvement on the reasoning scale will result in something that is equivalent enough to humans on the intelligence axis to replace all labor.

addaon• 2 months ago

https://xkcd.com/605/

inglor_cz5• 2 months ago

Hmmm, this is quite an interesting take by Scott.

Lindy's Law is not actually a law and many exact minds will be provoked by the very name; it also fails spectacularly in certain contexts (e.g. lifetime of a single organism, though not necessarily existence of entire species).

But at the same time, I am willing to take its invocation in the context of AI somewhat seriously. There is an international arms race with China, which has less compute, but more engineers and scientists. This sort of intellectual arms race does not exhaust itself easily.

A similar space race in the 1950s and 1960s progressed from first unmanned spaceflight to a moonwalk in mere 12 years, which is probably less than what it takes to approve a bicycle lane in Chicago now.

krupan3• 2 months ago

"There is an international arms race with China"

I keep seeing this. Where did it come from? Has China said that they intend to attack other countries using AI? Have other countries declared that they intend to attack China with AI?

Also, why does anyone believe that AI could actually be that dangerous, given it's inherent unpredictable and unreliable performance? I would be terrified to rely on AI in a life or death situation.

inglor_cz5• 2 months ago

It was a metaphor. I meant, and later clarified, an intellectual arms race.

BTW your handle is an actual Czech word, minus a diacritic sign ("křupan"), and a bit amusing one. It basically means hillbilly. Not that it matters, just FYI.

Anyway: AI will be used in military context, and it probably already is. Both for target acquisition and maybe even driving the weapon itself. As of now, the Ukrainians are almost certainly operating some AI-enabled killer drones.

dmbche• 2 months ago

https://www.forbes.com/sites/greatspeculations/2025/11/25/wh...

aspenmartin5• 2 months ago

AI in war is like Palintirs whole business model. You have a system that can effectively deal with ambiguity and has superhuman performance on reasoning plus superhuman physical abilities via embodiment…

Inherent unpredictable and unreliable performance is also quite the feature of human beings as well.

mitthrowaway24• 2 months ago

It's not a law per se, but there are rules for reasoning under uncertainty to get the most out of what limited knowledge you have, and Lindy's law arises from that. To do better than Lindy's law requires having additional information about the problem beyond just the one data point.

krupan3• 2 months ago

News flash: predicting the future is hard

energy1233• 2 months ago

The individual who is the best at predicting the future is predicting ASI and full labor automation by 2040:

https://xcancel.com/peterwildeford/status/202963666232244661...

gerikson• 2 months ago

Past results is no guarantee of future performance.

Aurornis• 2 months ago

The individual who is the best at predicting the future

Going to need a big citation for that claim

margalabargala2• 2 months ago

The individual who is the best at predicting the future

Lol

layer8• 2 months ago

Predicting who will predict the future best is hard.

dsign2• 2 months ago

My own bet is end of that decade: somewhere between 2045 and 2050.

Ofc "full labor automation" has a certain spread of meaning. A sliver of population will always find ways to hold to a job or run one or many businesses. But there will be "enough" labor automation for it to be a social ticking bomb. That, in fact, does not depend on better models nor better AI than we have today. By 2045 there will be a couple of generations that has been outsourcing their thinking to AI for most of their adult lives. Some of them may still work as legal flesh of sorts, but many won't get to be middle man and will find no job.

Also, if you could replace your senator today by an untainted version of a frontier model (of today), would you do it? Would it be a better ruler? What are the odds of you not wanting to push that button in the next twenty years, after a few more batches of incompetent and self-serving politicians?

solid_fuel3• 2 months ago

The individual who is the best at predicting the future

Yeah well my prophet says he can beat up your prophet in a fight.

---

Here in reality, I'm not accustomed to taking random predictions without backing evidence as if they were truth.

Alex_L_Wood3• 2 months ago

I can’t take anyone who puts probability on events like that seriously. How would you even explain the calculations behind that with a straight face?

FergusArgyll• 2 months ago

Wait till you read his piece about Fordow

kubb6• 2 months ago

If the scary AI is so inevitable, why do you feel such an overwhelming need to convince people about that? Surely you can just wait a bit, and they'll see for themselves.

adleyjulian2• 2 months ago

1. It's not inevitable. 2. Those that see AI as an existential risk don't generally think it's a guarantee, but if it's say a 5% chance then that's worth addressing/mitigating. 3. That's not what this article was even about.

kubb6• 2 months ago

Sounds like the burden is on you to explain either

  1. If you're not treating my claim as a black box, explain explicitly what is your model of what the article was about? Are you aware, for example of the last paragraph of the article? I think that WAS what the article was about. Do you have specific opinions on e.g. how I went wrong and where my model differs?
  2. If you are treating it as a black box, what's your default expectation based on the law of Nothing Ever Happens?

Just kidding, you don't need to explain anything. A"I" fearmongers should though.

mitthrowaway24• 2 months ago

By that reasoning, why even warn people about anything? Why do road construction crews put up signs saying "ROAD CLOSED AHEAD" when you can just drive on and see for yourself?

kubb6• 2 months ago

Indeed, why warn people about real things that exist in the world? That is EXACTLY the same as inciting fear about something imaginary (not even projected).

throwawayk7h3• 2 months ago

Yeah! And if climate change is so inevitable, why do the people who want to prevent it from happening seem hell-bent on convincing people that climate change is real?

LarsDu884• 2 months ago

I think an interesting thing about recent AI developments is that its all happening right as we hit the diminishing returns side of another "exponential that's actually a sigmoid" which is Moore's law.

The naive expectation is that AI will slow down b/c Moore's law is coming to an end, but if you really think about the models and how they are currently implemented in silicon, they are still inefficient as hell.

At some point someone will build a tensor processing chip that replaces all the digital matmuls with analogue logamp matmuls, or some breakthrough in memristors will start breaking down the barrier between memory and compute.

With the right level of research funding in hardware, the ceiling for AI can be very high.

cyanydeez2• 2 months ago

they already did put a model into the silicon and it's crazy fast. https://chatjimmy.ai/

I'm pretty sure there's a 3 year design goal starting this year that'll do that to any of the qwen, deepseek, etc models. There's a lot you could do with sped up models of these quality.

It might even be bad enough that the real bubble is how much we don't need giant data centers when 80-90% of use cases could just be a silicon chip with a model rather than as you say, bloated SOTA

clickety_clack• 2 months ago

It would be pretty cool to have interchangeable usb keys with models on them.

LarsDu884• 2 months ago

And this is an asic that is still operating digitally. Imagine a chip with baked it weights that does its math analogue with 20x reduction in number of circuit elements needed to do a multiplication op.

If there's a breakthrough in memristors, you could end up with another 20x reduction in circuit elements (get rid of memory bottlnecks, start doing multiplication ops as log transform voltage addition)

The ceiling is ultra high for how far AI can go.

throwaway274484• 2 months ago

Even at orders of magnitude greater speed, we've still hit diminishing returns for quality of output. We simply haven't found anything like superhuman reasoning ability, just superhuman (potentially) reasoning speed.

horsawlarway• 2 months ago

Possibly - but we've also seen that spending more tokens on a task can improve the quality of the output (reasoning, CoT, etc).

So it's not impossible to have things that seem orthogonal, like generation speed or context length, have an impact on quality of result.

energy1233• 2 months ago

It's not that easy to assess diminishing returns with saturated benchmarks where asymptoting to 100% is mathematically baked in. I could point to the number of Erdos proofs being solved by AI going from 0 to many very recently as evidence for acceleration.

LarsDu884• 2 months ago

I disagree with this. Reinforcement learning with verifiable rewards training is actually the secret sauce that is leading Claude and GPT to automating software engineering tasks.

All the easily verifiable domains such as mathematics, coding, and things that can be run inside a reasonable simulation are falling very very fast.

By next year if not sooner, mathematicians will be wildly outpaced by LLMs for reasoning.

paulpauper3• 2 months ago

Moore's law is bypassed with volume--more datacenters

torginus3• 2 months ago

The argument I've heard about how special Moore's law is that if you take a baby crawling at 10cm/s and 1000x it, you get almost Mach 3, which is almost as fast as the SR71 Blackbird flies. yet with a 18 month cadence, that's 15 years of progress, which has happened multiple times during the existence of the VLSI industry.

rdedev• 2 months ago

IMO we are either limited by data or reaching the limits of what's possible with a transformer architecture. Hardware will get us efficiency but I am not sure if it will lead to smarter models

dehrmann• 2 months ago

My doubts in the architecture is how different they are from human intelligence. They need an inordinate amount of training data and lack any sort of generational architectural intelligence.

ToValueFunfetti5• 2 months ago

I suspect if you consider neurons as components of computation, you could draw an exponential of total computation in the world that goes back to the dawn of humanity, maybe further. Most of that would just be population, but interesting that digital computers start picking up the slack just as population growth slows.

imtringued3• 2 months ago

but if you really think about the models and how they are currently implemented in silicon, they are still inefficient as hell.

I'm not sure this follows? Research has gone into two bit quantizations that only need a scale factor per block and each parameter merely takes up two bits which means that the operations can be mapped directly onto adders rather than multipliers.

but if you really think about the models and how they are currently implemented in silicon, they are still inefficient as hell.

The vast majority of analogue components are both space and energy inefficient. Digital won for a reason. You can simply keep scaling to lower voltages and smaller transistors since you only need to distingish between one and zero.

Brendinooo• 2 months ago

then what is their model?

My mental model has been 3D computer graphics: doubling the polygon count had huge returns early on but delivered diminishing returns over time.

Ultimately, you can't make something look more realistic than real.

I don't know what the future holds, but the answer to the question "can LLMs be more realistic than real" will determine much about whether or not you think the curve will level off soon.

the8472• 2 months ago

The equivalent bar in this domain would be human intelligence, and we already have growing lists of tasks where machines outperform humans. We even known of natural systems that outperform humans on some metrics, e.g. bird-brains have higher neuron density than ours because evolution had to optimize more for weight.

kilpikaarna• 2 months ago

In 3D graphics there's diminishing returns on investment for the technology itself, but the real limit is one of economics. How to create all those assets required by the rendering tech, and make your money back. Preferably while also keeping your customers interested long term, by not becoming risk averse.

In the same fashion, LLMs have to pay for themselves to keep the trendlines going. In a whole-systems -sense, mind, not "$2000/month is cheaper than hiring a developer" while the rest of the economy collapses.

btilly4• 2 months ago

Lindy’s Law is an absolute gem, that I'm keeping.

If we don't understand the fundamental limits to any particular kind of trend, our default assumption should be that it will continue for about as long as it has gone on already.

We can, in fact, easily put a confidence interval on this. With 90% odds we're not in the first 5% of the trend, or the last 5% of the trend. Therefore it will probably go on between 1/19th longer, and 19 times longer. With a median of as long as it has gone on so far.

This is deeply counterintuitive. When we expect something to last a finite time, every year it goes on, brings us a year closer to when it stops. But every year that it goes on properly brings the expectation that it will go on for a year longer still.

We're looking at a trend. We believe that it will be finite. Our intuition for that is that every year spent, is a year closer to the end. But our expectation becomes that every year spent, means that it will last yet another year more!

How can we apply that? A simple way is stocks. How long should we expect a rapidly growing company, to continue growing rapidly?

jerf• 2 months ago

It's an interesting idea, and it may be something that could be mathematically justified, but I do think this is an abuse of Lindy's Law in the absence of such a justification. Per Wikipedia [1]:

"The Lindy effect applies to non-perishable items, like books, those that do not have an "unavoidable expiration date"."

And later in the article you can see the mathematical formulation which says the law holds for things with a Pareto distribution [2]. I'd want to see some sort of good analysis that "the life span of exponential growth curves" is drawn from some Pareto distribution. I don't think it's completely out of the question. But I'm also nowhere near confident enough that it is a true statement to casually apply Lindy's Law to it.

[1] https://en.wikipedia.org/wiki/Lindy_effect

[2] https://en.wikipedia.org/wiki/Pareto_distribution

btilly4• 2 months ago

The analysis in the article explains why it applies to any phenomena that we might be able to notice.

The argument given is the same as the one that I first ran across, not by that name, in https://www.nature.com/articles/363315a0. https://en.wikipedia.org/wiki/Doomsday_argument claims that it was a rediscovery of something that was hypothesized a decade article.

I hadn't tried to give it a name, or thought to apply it outside of that context.

As for the mathematical qualms, I'm a big believer in not letting formal mathematical technicalities get in the way of adopting an effective heuristic. And the heuristic reasoning here is compelling enough that I would like to adopt it.

tomjakubowski• 2 months ago

People who correctly cite the Lindy effect won't look like people who correctly cite the Lindy effect.

riknos314• 2 months ago

those that do not have an "unavoidable expiration date"."

Try avoiding the heat death of the universe /s

LPisGood• 2 months ago

This is the exact same heuristic used in CPU scheduling.

We expect fresh processes to terminate quickly and long running processes to last for a while longer.

skybrian3• 2 months ago

You can do that but you're laundering ignorance into precise-seeming mathematics. Better to just say "we're probably somewhere in the middle, not at the beginning or end" and leave it at that. Calling a peak is hard.

btilly4• 2 months ago

You speak about laundering ignorance into precise-seeming mathematics as if it was a bad thing.

But that's the entire idea of Bayesian reasoning. Which has proven to be surprisingly effective in a wide range of domains.

I'm all for quantifying my ignorance, and using it as an outside view to help guide my expectations. Read the book Superforecasting to understand how effective forecasters use an outside view to adjust their inside view, to allow them to forecast things more precisely.

cortesoft• 2 months ago

I feel like Lindy's law doesn't work for things whose observation is partly controlled by the thing itself.

For example, take something like a fad or trend; they don't have a hard end date like human lifespan, so it should follow Lindy's law.

However, the likelihood, on average across the population, that you observe a trend is going to be higher at the end of a trend lifecycle than at the beginning. This is baked into the definition - more and more people hear about a trend over time, so the largest quantity of observers will be at the end of the lifecycle, when the popularity reaches its peak.

In other words, if you are a random person, finding out about a trend likely means it is near the end rather than the middle.

mike_hock• 2 months ago

Similarly, if you are a random person being alive, it likely means that the world population is near its peak and extinction is at hand, or at least the start of a permanent decline.

We have at least global warming and impending WW3, so that line of reasoning seems to work.

ccppurcell• 2 months ago

Well it only works when there is no information at all apart from the past frequency.

It's the solution to the tank problem. You know that the enemy number their tanks as they're produced. You capture a tank and know its number, N. What's the best guess about how many tanks the enemy has produced so far? As a pure mathematical model with no other details, the best guess is 2N. Of course in reality you have some ideas about how long it takes to make a tank, how many resources the enemy has etc.

Analogously you have information about the way trends develop.

tsimionescu3• 2 months ago

While this is very fun as a mathematical exercise, it's completely irrelevant as a real tool for getting a better understanding of unknown processes in the real world.

The law only applies for certain types of processes, and is completely wrong for other types (e.g. a human who has lived 50 years may live 50 more, but one who has lived 100 years will certainly not live 100 more). So the question becomes: what type of process are you looking at? And that turns out to be exactly the question you started with: is there a fundamental limit to this growth curve, or not.

jfjfnfnttbtg• 2 months ago

The law only applies for certain types of processes

Did you even read the post? It’s an estimate in the context where you have zero information on which to base an accurate estimate. The author’s point is that if you’re making a different estimate you need to actually say what information is informing that.

Human lifespan is obviously not a case where we have zero information, so what is your point in bringing that up?

dado3212• 2 months ago

But if you met an alien who said they'd been alive for 100 years you wouldn't assume they're on the verge of dropping dead: you would assume they live longer. It's a rough rule for when you don't have other information, and if you're arguing against it you need to specify what other information you're using to make that argument.

throwawayk7h3• 2 months ago

Closely related is Laplace's Rule of Succession[1], which basically says that (in lieu of other information), the odds of something happening next time go down the more times in a row that it doesn't happen (and vice versa).

So for example, the longer a time bomb ticks, the less likely it is to go off any time soon. (Assuming the timer isn't visible.) :)

[1] https://en.wikipedia.org/wiki/Rule_of_succession

coldtea5• 2 months ago

If we don't understand the fundamental limits to any particular kind of trend, our default assumption should be that it will continue for about as long as it has gone on already. We can, in fact, easily put a confidence interval on this. With 90% odds we're not in the first 5% of the trend, or the last 5% of the trend. Therefore it will probably go on between 1/19th longer, and 19 times longer. With a median of as long as it has gone on so far.

People would confidently cite Lindy's law all the way near the end of a trend. Nothing would stop a Roman saying that just before the Fall.

We don't always need to "understand the fundamental limits" to a trend to see where it's going. Just to observe more than a random blind guess about them.

I also wouldn't trust the "see how much we're improving" benchmarks of a trillion dollar pre-IPO industry to begin with.

robberat7• 2 months ago

I am sure Lindy's law is not a general law. Because i have a completely opposite example- How many years can we go by without having a passenger plane crash? as the number of years with no crash increases the chance of something going wrong increases with it. If a crash happened last week, you'd have some sort of sense that okay, it's not going to happen the week after. I can't comment on the causal mechanism, but maybe the collective of people working in the flight industry become more aware after a crash and do their job well.

itkovian_• 2 months ago

The other thing people don’t understand is exponential curves are self similar. The start of an exponential looks like an exponential. People always look at and think ‘well that’s it it’s exponential now, have missed it, can’t sustain’. Nope.

Good example of this is number of submissions to neurips/icml/iclr. In 2017 that curve was exponential.

zkmon• 2 months ago

The curve is a smoothed step curve (y=1 if x>1 otherwise 0). Nature doesn't allow any change to happen instantly at any degree of rate of change. The curveis just a manifestation a change with exponential smoothening of the sharp corners.

For example, When a car starts, it's speed and acceleration become more than zero. But what about rate of change in higher degrees? It suddenly doesn't change from zero acceleration to non-zero. That means the car has a non-zero derivative at all degrees. In other words, the movement is exponential. The same thing happens in reverse when the car reaches a constant speed.

patrickmay• 2 months ago

Stein's Law: "If something cannot go on forever, it will stop."

skybrian3• 2 months ago

Yes, but figuring out when is the hard part.

OscarCunningham• 2 months ago

John D Cook gives more technical details here: "Trying to fit a logistic curve" https://www.johndcook.com/blog/2025/12/20/fit-logistic-curve...

bedobi2• 2 months ago

[flagged]

simianparrot2• 2 months ago

Because HN is YCombinator which has invested in probably hundreds of «AI» firms by now. Including OpenAI.

Allowing slop articles like this literally prints them evaluation money.

t435625• 2 months ago

Yes, this is not the place to express skepticism of any kind.

ngriffiths• 2 months ago

I think there are many ways someone with his lack of expertise can still be valuable, including:

- Making connections to other subjects that an expert would miss. The hall of fame of sigmoid predictions is just excellent, I already know I'm going to be reminded of it some time in the future. Very entertaining way to get the point across.

- Writing about tricky concepts in a very accessible and elegant way, which experts are notoriously bad at doing themselves - they are often optimizing for other specialists.

- Being able to write with an air of speculation and experimentation with ideas that experts and institutions often can't afford. Experts have to maintain their track record; Scott Alexander can say "lol just double the timeline"

bedobi2• 2 months ago

[flagged]

tomhow2• 2 months ago

Please don't post like this on HN. The guidelines explicitly ask us not to sneer or be curmudgeonly - https://news.ycombinator.com/newsguidelines.html.

As for the basis of your objection, this smacks of intellectual gatekeeping. Plenty of good writing is by people who are not academically qualified or a recognized expert in the topic they're writing about. Indeed, very often, this kind of writing is better than writing by experts. Experts often write for other experts, and this can be exclusionary to lay readers. When a non-expert learns about a topic then writes about it for a general audience, they tend to be just a step ahead of the audience, and so the reader is able to learn about the topic by following the process of discovery and reasoning that the author just experienced. Sure, they often get some details or concepts wrong, but the discussion on a site like HN can draw other perspectives, and – very often – contributions from experts, which leads to further expansion in everyone's understanding of the topic.

HN's very ethos is to gratify intellectual curiosity, and this kind of writing is highly compatible with that.

janalsncm• 2 months ago

What if you don’t fully understand the process? AI forecasters know some things (like how data centers work and how much it costs to build them). But they’re unsure about other things (researchers keep inventing new paradigms of data generation that get over data walls, but for how long?), and other things are entirely opaque (What is intelligence really? Why do scaling laws work? Might they just stop working at some point?) Is there anything you can do here?

This is the crux of the article. To a large extent continued progress depends on a stable increase in compute, an increase in training data, and an increase in good ideas to squeeze more out of both of them.

One calculation you could do is a survival function: for each of the above, how long before it is disrupted? For example, China could crack down on AI or invade Taiwan. Or data centers become politically unpopular in the US. Or, we could run out of great ideas. Very hard to predict.

dsign2• 2 months ago

We did hit the sigmoid's plateau on airplane speed, but the applications of airplane speed are still coming (how fast can a Chinese company airship the PCB you ordered three minutes ago?). I expect the the same will happen with LLMs, though I also happen to believe things are just getting started on end capabilities.

jsmcgd• 2 months ago

It’s true that birth rates must eventually flatten out and become sigmoid

All positive growth eventually flattens out and becomes sigmoid, but a lot of phenomena experience negative growth and nose dive. No gentle curve, but a hard kink and perfect flat line at zero. Forever. I think it would be a stretch to categorize that pattern as sigmoid. Predicting a sigmoid pattern for negative growth implies some sort of a soft landing (depending on your definition of soft).

We can think of many populations that are no longer with us. So just a caution about over applying this reasoning in the negative case.

Qem• 2 months ago

All positive growth eventually flattens out and becomes sigmoid, but a lot of phenomena experience negative growth and nose dive.

https://en.wikipedia.org/wiki/Seneca_effect

ngruhn• 2 months ago

all exponentials eventually become sigmoids

Except innovation. When one sigmoid tapers off we keep finding new ones to keep the climb going.

stymaar• 2 months ago

I don't know when the sigmoid is going to kick in, but Nvidia's Quaterly datacenters revenues have been grown 15 folds over the past 3 years[1], and nobody including Scott believes this is sustainable for 3 more years otherwise Nvidia's market cap would conservatively be at least an order of magnitude higher than it is.

All exponential eventually becomes a sigmoid because exponential growth always expose limiting factors that weren't limiting at the beginning. Silicon manufacturing had lots of room for high-margin customers like Nvidia even a year ago (by the mere virtue of outbidding lower-margin customers), but now it is mostly gone, and no amount of money will make fabs build themselves overnight.

[1] https://stockanalysis.com/stocks/nvda/metrics/revenue-by-seg...

pyrale• 2 months ago

Such a long article to say that neither side has a fucking idea about what will happen next.

While we're at it, the "exponentials are actually sigmoïds" meme is not necessarily true. While exponentials are never exponentials, sigmoids are not guaranteed. Overshoot-and-collapse examples also happen in tech, e.g. the dotcom bubble, or the successive AI winters.

andrewflnr• 2 months ago

It's really not that long, and is quite clear that its main point is about how to reason when you realize no one actually knows what's going on.

dreambuffer3• 2 months ago

FYI: The author has predicted that "AGI" will be here in 1-2 years and has staked his public reputation on it. He is personally invested in trendlines being lindy rather than sigmoid.

I don't think you can use lindy on trends as if trends are static objects, but that's another conversation.

throwawayk7h3• 2 months ago

Mind you, he is only personally invested insofar as he's staked his reputation on it. Throughout his writing, he expresses the same point over and over again: desperately wants AI to slow down, advocates for politics that would slow it down, and most likely nothing would bring him greater peace than to see a sigmoid curve appear.

bombcar• 2 months ago

How convenient; when AGI doesn’t appear in 1-2 years his reputation is pristine because he slowed it down.

woeirua2• 2 months ago

Ok, but you can just look at the METR curve. Mythos saturated the 50% time horizon. The 80% is now at 3 hours. The rate of progress is accelerating not slowing down. There’s no indication yet that this is a sigmoid!

yorwba2• 2 months ago

The METR task set contains no tasks with a duration greater than 32 hours (conservatively eyeballed from Figure 3: https://arxiv.org/abs/2503.17354 ), so any prediction that naively forecasts a longer time horizon is trivially incorrect. I guess that won't lead to a sigmoid-looking graph though, since METR will likely switch to a different evaluation methodology at that point and stop updating the old curve.

paulpauper3• 2 months ago

He only has 1.5 more months. If he's wrong he needs to own it. Same for Eliezer Yudkowsky. But these people have too much riding on their brands. No one has the courage to fess up to being wrong. Given how many podcasts he and others have been on professing this belief, it will be hard to just pretend otherwise.

quickthrowman• 2 months ago

Eliezer Yudkowsky, now there’s a name I haven’t heard in a while.

What has he been up to since finishing the finest work of literature ever produced, Harry Potter and the Methods of Rationality? I’ve been patiently awaiting a sequel!

hollerith• 2 months ago

Yudkowsky has never predicted that "AGI" will be here in 1-2 years. He has been saying frequently for years that it is easier to predict how the AI juggernaut will turn out (i.e., very badly for us) than to predict when the very bad things will happen.

(I don't know about the other guy mentioned above.)

Sniffnoy4• 2 months ago

FYI: The author has predicted that "AGI" will be here in 1-2 years and has staked his public reputation on it. He is personally invested in trendlines being lindy rather than sigmoid.

I mean, that's called "having an opinion".

paulpauper3• 2 months ago

He wrote articles arguing that pro-AI people are dismissive of risks or even suggesting they are intellectually lazy. He's taken a side. if he's wrong I would hope he owns up to it

dreambuffer3• 2 months ago

He co-authored a report, which is something more than an opinion. It may be used to inspire policy. There should be greater reputational consequences for publishing something you spent a few months studying and writing about along with several experts. Just my opinion.

DonsDiscountGas• 2 months ago

And now he's publishing more information about that same opinion he still has. How horrible.

sigmoid102• 2 months ago

AGI has become such a meaningless nondescript term, arguing when or how it is here has become pointless. Even OpenAI caved in and removed their AGI clause from their contract with Microsoft because they weren't fully sure that we are not there yet. The original ARC AGI was hailed as proof that AGI is not here yet, but now that ARC 1 and 2 got saturated, noone wanted to consider that perhaps we crossed the point where average humans are getting left behind. Frontier models are primarily limited by context and modality at this point, not by intelligence.

Auracle• 2 months ago

To your point, if we had truly unlimited context to the point where at least that instance of a model could “learn” and have what seems like a continuous “consciousness” I think many of us would think that we’ve attained AGI.

Right now we have an incredibly smart thing with severe short term memory loss, and it’s hard for us to reconcile that as it’s so different from us.

ToValueFunfetti5• 2 months ago

This is incorrect as written. The author contributed writing to AI-2027 but distanced himself from the underlying model. That model had 2027 as the modal year of AGI, not median or mean. The authors of that model revised it to a later date shortly after and (if I recall correctly) have since done so again.

It is broadly true that Scott believes that AGI will come in the near future and from LLMs, although his reputation runs a ways deeper than that.

dannyobrien• 2 months ago

So, this is not quite right: Alexander contributed to the report, but his personal opinion is more like the mid-2030s[1]. Freddie feels like this is him backing down from the original statement, but in fact he said this at the time the report was published, and in fact pointed out a graf below the quote that Freddie claims does tie him to 2027:

Do we really think things will move this fast? Sort of no - between the beginning of the project last summer and the present, Daniel’s median for the intelligence explosion shifted from 2027 to 2028. We keep the scenario centered around 2027 because it’s still his modal prediction (and because it would be annoying to change). Other members of the team (including me) have medians later in the 2020s or early 2030s, and also think automation will progress more slowly. So maybe think of this as a vision of what an 80th percentile fast scenario looks like - not our precise median, but also not something we feel safe ruling out. [2]

I don't think this changes your observation that he is "personally invested" (i.e. believes this trendline will continue), but I'm pretty sure when AGI doesn't appear in 2027, many people will believe that this invalidates the arguments being made here (or in the report). The actual report was intended to give a feel for what a near-future "disaster" AGI scenario, and settled on a date to give that some concrete immediacy. The collective review that gave that as a possible, but not inevitable date is still ongoing (they originally pushed their best estimate out a bit further, but now they think, judging by the goals that are being hit, their scenario was a little too conservative). [3]

[1] https://freddiedeboer.substack.com/p/im-offering-scott-alexa... [2] https://www.astralcodexten.com/p/introducing-ai-2027 [3] https://blog.aifutures.org/p/grading-ai-2027s-2025-predictio...

grey-area3• 2 months ago

AI boosters really are detached from reality.

LLMs are nothing close to AGI and not going to lead to it, they can’t distinguish right from wrong, they can’t count, they can’t reason, they generate plausible text from a vast databank of connected text.

Apparently that is enough to fool many people but it’s nothing close to AGI which would require internal models of the world, reasoning etc.

We are nowhere close to AGI and the fools who predicted we were will unfortunately keep lying about their stated timelines when it inevitably doesn’t arrive. You’re already hedging and trying to caveat previous predictions, as OpenAI did with their AGI predictions which they’re now furiously back-pedalling on.

scott012• 2 months ago

If I'm not mistaken, he's either affiliated with or otherwise connected to the effective altruist movement, hence he can't be unbiased. I find this article tells an interesting perspective on it: https://www.noemamag.com/the-politics-of-superintelligence/

naasking7• 2 months ago

Nobody's unbiased.

goosejuice• 2 months ago

The whole rationalist movement is super bizarre to the "normies".

It's not at all surprising that they are increasingly getting labeled a cult (they aren't by traditional definition but there are a lot similarities). I'm really surprised it hasn't hit the mainstream yet given the connections to Elon, Thiel, frontier labs, dark crypto funding, FTX/SBF, some suicides and some murders. It's all a little nuts.

Meanwhile you got all the anti-democratic NRx people on the other side of it.

I suspect this new doc coming out on HBO will spark a media frenzy.

andy99• 2 months ago

AI has scaled well according to convenient measures. It (neural networks) have the property that whatever you define, they can rapidly be trained master it. We’re able to show that various tasks of increasing complication do not require intelligence and can be framed as autoregressive RL problems. I personally don’t think AI is any closer to sentient intelligence than LeNet; it’s almost trivially clear, we know how it works. So we’re measuring something orthogonal, basically how well a universal function approximator can fit to a function we define, given arbitrary computing power, and calling that progress. What will be really interesting is if we’re able to find a way to properly measure what they can’t do and what’s different about real intelligence.

Edit: in particular I don’t agree with

  But if someone claims that the trend toward increasing AI capabilities will never reach some particular scary level...

One has to agree that the benchmark results are getting “scarier”, which is not automatically implied by finding more goals to optimize for

ordu• 2 months ago

We’re able to show that various tasks of increasing complication do not require intelligence and can be framed as autoregressive RL problems.

The important thing we can show it in hindsight only. We don't know which other tasks we are currently mistaken about requiring intelligence. Maybe none of them are?

We don't know. We don't know what intelligence is. If we look at decades and even centuries of attempts to define intelligence, it is all looks like a goalposts moving. When a definition of intelligence starts to include people or things we don't like to think as of intelligent ones, we change the definition.

ryeights2• 2 months ago

“AI is whatever hasn't been done yet” — Larry Tesler

winwang• 2 months ago

Yep. No one bats an eye at eyewitnesses "hallucinating" details, or that I'd rather have Opus as a coworker vs a random middle schooler (err, labor laws notwithstanding). I think perhaps too much of the dialogue around intelligence has to do with the word (and its connotations) itself.

The poster you replied to even used the word "sentient", which is quite interesting (warning: opinionated tangent ahead). Merriam-Webster defines it as "capable of sensing or feeling: conscious of or responsive to the sensations of seeing, hearing, feeling, tasting, or smelling". Feels like qualia. Or if we don't want to go the qualia route... Of course, we wouldn't call Helen Keller non-sentient, so presumably we "really" mean "can it sense or feel" -- well, sense is just "act/feel according to the environment", which you could argue in the case of an LLM would be their context... so we should "really" remove "sense" from the definition, probably. So "do LLMs feel" is probably closer to what "sentient" is being used for here. Since we don't have the obvious symmetry of "you are like me and I feel (therefore you probably feel)", it's way better/easier/feel-good-ier to prefer "LLMs don't feel" rather than "oh shit, it feels and model training is actually just torturing it into the right shape". LLMs as fundamentally non-intelligent also avoids the problems of "what does that say about people" or "we may have made 'AGI' and it wasn't what we thought it would be" or "we're not ready to talk about this yet".

red75prime2• 2 months ago

basically how well a universal function approximator can fit to a function we define

That's what you've got wrong. We don't define functions that an LLM approximates. Autoregressive pretraining approximates an unknown function that produces text (that is what the brain does). RL doesn't approximate functions, it optimizes objective by finding an unknown function that performs better.

imtringued3• 2 months ago

What an interesting perspective! There is just a small problem. How does the human brain work in the phase between being born and learning written language? Seems like a significant bootstrapping problem for your theory.

jbjbjbjb• 2 months ago

Is it not the case that a lot of the recent gains are just the coding harness that directs the LLM? That coding harness isn’t all that intelligent, simple pattern matching that maps to well defined tasks a programmer might do.

naasking7• 2 months ago

I wouldn't say "just", but harnesses are a big deal responsible for a lot of improvements, yes. Models are also fine-tuned on these, like MetaClaw.

baxtr• 2 months ago

The moral of the story is that, even though all exponentials eventually become sigmoids, this doesn’t necessarily happen at the exact moment you’re doing your analysis. Sometimes they stay exponential for much longer than that!

All exponentials eventually become sigmoids? Don’t think this can be true without qualifiers.

jvanderbot• 2 months ago

All models are wrong, of course, but this is kind of "common sense" so it's not hard to accept as true in a natural system. How can something continue on exponential growth forever without reaching a new blocker that causes slowdown or encountering pushback that makes it an oscillator. A pendulum looks exponential when it is at its peak and accelerating down.

The issue is that the exponential-looking part of the sigmoid might contain all of human history, sure, but most folks who espouse this theory probably agree that over time everything reaches a steady-enough state to be considered non-exponential, or become oscillatory.

jrflowers• 2 months ago

I like this article about how we should assume, at any given point, that we are exactly halfway through a phenomenon which relies on a single data point on a graph —-that apparently doesn’t need its relevance or importance explained— to illustrate that this is obviously true for AI in particular

Smaug1234• 2 months ago

I think you may have read a different article from me. The thesis of the article is summarised at the end:

But if someone claims that the trend toward [X] will never reach some particular scary level, then the burden is on them to explain either:
If they’re not treating [X] as a black box, and claim to be modeling the dynamics explicitly, then what is their model? Have they calculated the obvious things…
If they are treating [X] as a black box, why isn’t their default expectation based on Lindy’s Law?

Like, the whole point is that in real life we do actually know things about situations and can model them; we fall back to Lindy's law when we know nothing at all. Further, arguments have justification to deviate from Lindy only when they give specifics about the situation they're modelling.

noosphr3• 2 months ago

This article answers the question in the second paragraph then completely ignores the answer for the rest of it.

My understanding is that this represents 3-4 “generations” of different technology (propellers, turbojets, etc). Each technology went through normal iterative improvement, then, when it reached its fundamental limits, got replaced by a better technology. The last technology, ramjets, reached its limit at about 3500 km/h, and there wasn’t the economic/regulatory will to develop anything better, so the record stands.

You don't have one sigmoid, you have multiple each stacked on top of each other. Airplanes aren't just one technology they are multiple technologies that happen to do the same thing.

Each one is following a sigmoid perfectly. It only looks exponential(ish) because of unpredictable discoveries that let you switch to another sigmoid that has a higher maximum potential.

The same is true in AI. If you used the same architecture as GPT2 today you're in for a bad time training a new frontier model. It's only because we have dozens of breakthroughs that the capabilities of models have improved as much as they have.

That said exponential and sigmoids are the wrong model to use for growth. Growth is a differential equation. It has independent inputs, it has outputs and some of those outputs are dependent inputs again through causal chains of arbitrary complexity. What happens depends entirely on what the specific DE that governs the given technology is. We can easily have a chaotic system with completely random booms and busts which have no deep fundamental rhyme or reason. We currently call that the economy.

Sniffnoy4• 2 months ago

Yes, I was surprised he never discussed the idea that such exponentials are typically made of stacked sigmoids.

That said... if the exponential is made of stacked sigmoids, it's still an exponential on the whole! The fact that it's made of stacked sigmoids is relevant to the engineers making it, but not so relevant to the users or those otherwise affected by it.

noosphr3• 2 months ago

Only so long as you can keep inventing the next sigmoid in the stack.

mediaman• 2 months ago

The book "Origins of Efficiency" by Brian Potter discusses this. Stacked sigmoids are a well-understood idea in innovation.

The idea that exponential growth will continue with stacked sigmoids is also not a given. An example is the nail. Nails used to be about half a percent of US GDP. That's a pretty big number! A series of innovations stacked on each other (each innovation having its own sigmoid) to reduce the cost of nails. Nails dropped in cost by over 90%.

But eventually nail manufacturing reached a floor. And since the mid-20th century, we haven't gotten much better at making nails. The cost of nails actually started increasing slightly. We ran out of new innovation sigmoids, so we got stuck on the last one.

So what you actually have to predict is whether there will continue to be new sigmoids, not whether the existing sigmoid will asymptote (we already know it will).

This is much more difficult to forecast, because new sigmoids (major new innovations) tend to be unpredictable events. Not only are the particulars difficult to forecast (if they were knowable, the innovation would have already happened), but whether there will be a major innovation or not is also hard to forecast, because they are distinct and separate from any existing sigmoid trend.

So we are left with the idea that all current innovations in AI will asymptote in their scaling as they reach the plateau of the sigmoid, but there may be new sigmoids that keep the overall trend up. Or there may not be. We don't know.

That's not very satisfying, so we'll get to keep reading articles like this one.

wcarss• 2 months ago

I don't disagree with you, but your example of nails and their cost reductions made me wonder whether we reached a meaningful limit in say, some fundamental material terms, or whether we just reached a limit in terms of return on investment.

Return on investment can be too low because the investment required is really high, but it can also be too low because the returns are just limited. If prices had dropped 90%, surely nails became even more ubiquitous, but at that stage there's only so much more money to dig out of the cost reduction hole. It feels plausible that there may have been ideas about more digging that could be done, but the reward just wasn't there in the market, especially versus just selling what worked.

I bring it up because the distinction in one specimen may speak to a larger trend: do new sigmoid developments tend to fail to materialize more often because of serious physical limits / lack of good ideas, or because of limitations to ROI? (Or, other things?)

In the arena of AI, the ROI on more intelligence/unit-cost seems pretty high right now. So, it seems like the difficulty of applying any potential innovations would have to be staggering for none to be pursued. Or, there'd have to just not be any good ideas to try.

Overall, I think there's ideas to try. So in my opinion, that shapes out to justify a bullish sentiment on sigmoids continuing to stack until the perceived potential gains from more intelligence/unit-cost somehow fall off.

Like I said, I don't disagree, we really don't know. But I feel it's a good bet that there's more coming.

Scene_Cast2• 2 months ago

Something that deeply frustrates me, as someone who did R&D on model architectures, is how similar the modern LLM model architectures are to GPT2.

(This is a bit disingenuous, as lots/most of work is spent on the scaling and training side of things.)

rictic• 2 months ago

The author has written about stacked sigmoids in the past, e.g. in https://slatestarcodex.com/2018/10/15/is-science-slowing-dow...

I suspect that he started down the path of considering them more deeply here but found that they didn't add much to the analysis. A stack of sigmoids ultimately either gives you a single sigmoid if you run out of new innovations, or an exponential if you don't.

loveparade• 2 months ago

If you used the same architecture as GPT2 today you're in for a bad time training a new frontier model. It's only because we have dozens of breakthroughs

What exactly are these dozens of breakthroughs? Most frontier models architectures today still look very much like GPT2 at their core. There were various of improvements like instructgpt, finetuning techniques, efficiency improvements with kv caches, faster attention, lora, better tokenizers, etc. Most of these are for making things run faster. The biggest differentiator has probably been data curation and post-training data and the ability to fit more into the model. But I think we had few breakthroughs that would fall into the category of different technologies.

pron2• 2 months ago

1. Scott Alexander is famous for writing about topics he knows little about. I'm glad to see he's found a subject he knows little about but so does everyone else.

2. What's even worse than predicting that some growth curve flattens before X happens is predicting it will flatten before X happens but after Y happens, which is what we see when it comes to AI in software development. Too many people predict that AI will be able to effectively write most software, replacing software engineers, yet not be able to replace the people who originate the ideas for the software or the people who use them. I see no reason why AI capability growth should stop after the point it's able to write air-traffic control or medical diagnosis software yet before the point where it's able to replace air traffic controllers and doctors.

3. While we don't know much about AI (or, indeed, intelligence in general), we do know something about computational complexity. Some predictions about "scary things" happening (the ones I'm guessing Alexander is alluding to, though I can't be certain) do hit known computational complexity limits. Most systems affecting people are nonlinear (from weather to the economy). Predicting them requires not intelligence but computational resources. Controlling them, similarly, requires not intelligence but either computational resources or other resources. It's possible that people choose to give control over resources to computers (although probably not enough to answer many tough, important questions), although given how some countries choose to give control to people with below-average intelligence (looking at you, America), I don't see why super-human intelligence (if such a thing even exists) would be, in itself, exceptionally risky.

ryeights2• 2 months ago

1. Scott Alexander is famous for writing about topics he knows little about. I'm glad to see he's found a subject he knows little about but so does everyone else.

This is kinda laughable. Scott has been thinking and writing about AI for a long time

pron2• 2 months ago

That's what I'm saying. I'm glad to see that he's found a subject where his lack of knowledge isn't a glaring handicap. When I come across his posts I usually feel uncomfortable because they read like a bright 4-year-old child trying hard to explain how a car works, only not cute (yeah, I know that trying to see what conclusions you can come to, Aristotle-like, from a basis of ignorance and without careful study is the whole point, but I never found this Memento-style game appealing).

leoc• 2 months ago

Hmm. What’s the general belief about Toby Ord’s “Are the Costs of AI Agents Also Rising Exponentially?” https://www.tobyord.com/writing/hourly-costs-for-ai-agents among those who are well-equipped to judge? Is it seen as wrong or disproven or unlikely? Because if not—if indeed recent LLM capability advances have likely relied on increases in inference cost per run which can’t be much further sustained—then it seems remiss not to mention that if you point to those advances to claim that the exponential trend remains on track.

lovich• 2 months ago

I wonder how the graph would look like if cost and/or profitability was taken into account.

I could probably make increasingly larger fires for years if I was willing to burn the entire world.

whatshisface2• 2 months ago

If you want a model, here's one: LLMs have never demonstrated the ability to go obviously beyond interpolating their training data. It takes an army of paid data producers solving homework problems to give ChatGPT the ability to do your homework. All vibecoded apps that turned out to be successful could put on a geological soil chart with other apps, probably on GitHub somewhere, on the corners. The prediction? They won't.

In this model, the exponential growth that everybody is freaking out about is only the realization of the modular software dream ("we'll only have to write an ORM once for all of human history!") and the sheer amount of knowledge in libraries.

It's at least falsifiable.

boxed7• 2 months ago

Just to play devils advocate: are we sure humans have demonstrated the ability to go beyond their training data? Like.. are we sure-sure about that?

1attice2• 2 months ago

Idk, I mean, Shakespeare never read Shakespeare, so, I mean, unless aliens?

whatshisface2• 2 months ago

I'm not asking for anything close enough to the boundary for questions like that to be difficult. There are some ML systems like AlphaGo that have crossed the line in specific domains. It's just that making self-play and online learning work for huge LLMs is highly non obvious.

The idea is simply that the basic idea behind LLMs, that you're distilling the entropy out of the entire available world of text, is antithetical to creativity.

Further developing on the theme of self-play, humans have the ability to sense what we want (intellectually) and reach for it communally over thousands of years. It's an innate quality, and if AI starts participating (contrast to giving people psychosis) we will all be able to tell.

suddenlybananas• 2 months ago

Planes, trains, and automobiles are not natural features of the environment.

rotis• 2 months ago

Are you suggesting intelligent design got us here?

ogogmad• 2 months ago

See the recent breakthrough "Erdos Problem 1196" which experts couldn't solve for 60 years until ChatGPT Pro did. ChatGPT's key idea was the use of the "Von Mangolt function" which it showed could finally settle the problem. Terry Tao has condensed the AI's proof to around a page. The problem was well-known to experts (in the field of number theory) but it was a Large Language Model that ultimately solved it - which it did without human help!

stego-tech2• 2 months ago

I felt the better takeaway from this was that it's impossible to know for certainty how long this will or will not continue regardless of the data or models you're using, because if you (or anyone else) could predict that accurately they'd be one of the richest people on the planet.

I don't know when (or if) AI will implode or succeed with any degree of provable certainty, because that's not my area of expertise. Rather, I can point out and discuss flaws in the common booster and doomer arguments, and identify problems neither side seems willing to discuss. That brings me cold comfort, but it's not enough to stake my money on one direction or another with any degree of certainty - thus I limit my exposure to specific companies, and target indices or funds that will see uplift if things go well, or minimize losses if things go pear-shaped.

I also think relying on such mathematics to justify a position in the first place is kind of silly, especially for technical people. Mathematical models work until they don't, at which point entirely new models must be designed to capture our new knowledge. On the other hand, logical arguments are more readily adapted to new data, and represent critical, rather than mathematical, thinking and reasoning.

Saying AI is going boom/bust because of sigmoids or Lindy's Law or whathaveyou is not an argument, it's an excuse. The real argument is why those things may or may not emerge, and how do we address their consequences within areas inside and outside of AI through regulation, innovation, or policy.

bjackman• 2 months ago

I think his agenda here is to point out that your probability distribution for AI outcomes should be broad (what you said), but most importantly: this means you must take seriously the possibility that we are gonna get superintelligence quite soon.

Basically a lot of people say "but isn't it also pretty likely that we DON'T get superintelligence?" And, yes, it is. But superintelligence being even a remotely plausible outcome is a big fucking deal. Your investment choices in that context are not important.

People really struggle to think rationally in the face of this shape of uncertainty.

stego-tech2• 2 months ago

You’re 100% correct, which is why I opted for a broad investment approach rather than trying to pick “winners”.

My thought process RE: superintelligence/AGI is generally this:

* I personally don’t believe it’s likely to happen with silicon-based computing due to the immense power and resource costs involved just to get to where we are now; hence why I invest broadly to capitalize on what gains we actually attain using this current branch of AI research across all possible sectors and exposure rates

* If we do achieve AGI using silicon-based computing, its limited scale (requiring vast amounts of compute only deliverable via city-scale data centers) will limit its broader utility until more optimizations can be achieved or a superior compute platform delivered that improves access and dramatically lowers cost; again, investing broadly covers a general uplift rather than hoping for a specific winner

* If AGI is achieved, nobody - doomer or booster alike - will know what comes next other than complete and total destruction of existing societal structures or institutions. The stock market won’t explode with growth so much as immediately collapse from the disintegration of the consumptive base as a result of AGI quite literally annihilating a planet’s worth of jobs and associated business transactions. In this case, a broad spread protects me from harm by spreading the risk around; AGI will annihilate the market globally, but not all at once barring a significant global catastrophe instigated by it

* Which brings me to the worst outcome, where AGI follows the “if anybody builds it everyone dies” thought process: investment is irrelevant because we’re all fucked anyway.

And that’s just my investment approach. I’m too pragmatic to believe we’re at the bottom of the sigmoid curve, but too wise to begin guessing where we actually exist on it at present or how much is left in the current LLM-arm of AI research; I’m an IT dinosaur, not an AI scientist.

What I can point to is the continued demand destruction of consumer compute through higher costs and limited availability due to rampant AI speculation as proof that the harm is already here in a manner most weren’t predicting, while at the same time actual job displacement by AI is limited to the empty boasting of executives using it as a smoke screen for layoffs after RTO mandates failed to thin headcount sufficiently.

In the USA in particular, we’re facing a perfect storm of:

* consumer confidence collapse leading to a decline in spending on all goods, especially luxury ones, by all but the most monied demographics

* data center-driven cost increases (energy) and resource destruction (land, water, fossil fuel use)

* the eradication of government support for renewable energy that would’ve kept these costs in check

* the widening wealth gaps creating a new underclass not seen since before WW2

In other words, most of the discourse continues to revolve around hypotheticals of tomorrow rather than realities of today. That would be the lesson I’d hope more people take away from something like this, so we can finally begin addressing issues themselves rather than empty online circle jerking about who is right or wrong.

XorNot3• 2 months ago

That's literally the singularity though - the point past which predictions are meaningless.

My "plan" is hope for a benevolent intelligence that establishes a post-human government and then enjoy poat-scarcity society doing wood working or something.

Billionaires should probably be more worried.

torginus3• 2 months ago

You want to go to the store to get ice cream. Ice cream is delicious and the value of eating ice cream is a small positive, let's say x. There's a one in ten million chance you'll get hit by a car on the way and die, and your life is infinitely precious, therefore the expected value of going is x times 1 = x, and the one of not going is 1/10m times negative infinity which is negative infinity. You are a rational person, so you don't go. In fact you don't do much of anything. Your value model of every activity has collapsed to a single value.

That's the problem with 'singularity' arguments. The people making them ignore the fact that the mathematical definition of the word means 'the model of outcomes collapses to a single value' therefore the model stops being useful, yet they somehow claim to be able to make predictions beyond the singularity. It's like those shitty Facebook math posts where they divide both sides of the equation by 0 (the fact hidden by some sleight of hand), to 'prove' that 2=1.

The formulation of the singularity involves putting outrageous values into the parameters of the model of reality, and denominator ignorance, and then claiming 'rationally' determining that the consequences are too severe to ignore.

topherhunt• 2 months ago

I think his agenda / point is that, viewed from Lindy's Law, given the SOTA in 2026, superintelligent AI arriving soon is vastly more probable than not, right? To make the case that "sure, AI capability and intelligence have grown exponentially over the past several years, but don't worry, they're about to abruptly level off and in fact won't blatantly surpass human-level intelligence within the coming decades" seems to have a high burden of proof unless your model is less "sigmoid" and more "abrupt plateau".

coldtea5• 2 months ago

but most importantly: this means you must take seriously the possibility that we are gonna get superintelligence quite soon.

So, his point with all the demand for rigor is to end on a hand-waved jump of faith from "improved AI models" to the mythical "superintelligence"?

par19702• 2 months ago

Which doomer argument have you found what problem with?

cellis• 2 months ago

Not to argue your broader point, but how, exactly would predicting this, in a macro sense, make you the “richest person on the planet”? A very uninformed person could tell you that “Cerebras will be the next NVIDIA”, put their life savings into it and in 10 years have 100 million. Thats not going to make them anywhere close to the richest person on the planet. Unless you can do it on a micro sense, merely having the ability to “predict” the way the wind is blowing is meaningless.

dnnddidiej• 2 months ago

Tldr you cant accurately predict the future of complex systems. Corollary... you cant accurately predict the future of complex systems using sigmoid.

Attention is all you need took us by surprise and we don't know how big the wave is let alone if there are other waves behind it.

graphememes• 2 months ago

line can go up, line can go sideways, line can go up sideways, line can go up sideways up, line go where line go

jasongi• 2 months ago

First sigmoid was transformers allow us to rapidly scale to our already abundant data until we tapped it out, the second is/was reasoning, allowing us to scale to our available compute (and compute manufacturing capacity). Correct me if I'm wrong but we don't have candidate for the third sigmoid, and scaling inference is hitting real-world supply chain constraints - electricity and chips.

Short of a third sigmoid appearing in the ML CompSci space, perhaps in the form of ongoing, repeated step-optimisations which will also have diminishing returns, intelligence growth is now limited a few scaling problems that have already been worked on for a very long time.

Transistors, which have been doubling for almost a century now, but Moores Law has already plateaued and reached limits on energy efficiency, and simply building new fabs is not something that we can do exponentially. And the other growth limiter is electricity - there is no exponential supply of fossil fuels or power plants. Although manufacturing has scaled, PV tech improvements are also plateauing - and while storage is getting cheaper, it's still not economical vs fossil fuels (meaning: when we have to switch to it, the growth slows down further) and we are unlikely to see battery efficiency sigmoid enough to maintain the AI sigmoid.

I don't mean to be bearish here. There's so much money sloshing around that we can afford to put the smartest people, using unlimited tokens, on the task of finding small, incremental gains on the CompSci side of things that will have large monetary payoffs - hopefully allowing further scaling and increased emergent abilities of LLMs. Maybe we can squeeze the algos for quite a while. But I don't see that maintaining the same level of exponential as unlocking unlimited data or maxxing out the world's energy/fab capacity for long.

And I don't see why this is a massive issue except for the people who want to have some god-like super AI? Frontier LLMs are genuinely magic. Not "won't delete your production database" magic, but definitely a massive productivity gain for competent knowledge workers.

foobar10000• 2 months ago

We are multiple orders of magnitude away from Landauer limits - so next big thing in matmul could be photonic multipliers - there’s a bunch of them coming up in the next 3? years. So that’s a 2-4 order of magnitude improvement. Sigmoid?

ianm218• 2 months ago

Reinforcement learning has become a huge portion of compute used during training runs [1] and synthetic data is letting us get lots more mileage out of the existing data. Additionally, there is lots of new, high quality data being created and collected each day. I think the "running out of data" thing was pretty poorly reported by mainstream media.

[1]. https://www.dwarkesh.com/p/dario-amodei-2

mgraczyk• 2 months ago

Don't bet on the sigmoid flattening out any time soon. I don't know when it will flatten, but it definitely won't be within the next 2 years

overgard• 2 months ago

This feels like a really verbose way of saying "things have been growing fast for a while so they should continue to grow as fast for just as long", and then he places the burden on people to prove him wrong. Um, no, the burden of proof is shared, "this will just keep going" requires just as much proof as "this is going to level off" if you're just looking at trend lines.

It's better to look at the underlying factors. Money sources are drying up, nobody is making a profit outside of nVidia, most blackwell GPU's are likely not even installed yet and will probably be 2 generations behind when they finally are being used, data centers are hitting all sorts of obstacles getting built and powered and they're getting built slowly, most AI researchers seem to think that LLMs are a dead end, the newer models seem to be getting more expensive and sometimes worse, or even potentially are showing signs of model collapse (goblins..), the supposed productivity gains are not materializing.. AI has worse public sentiment than congress.. I could keep going. Some obscure "law" seems to pale in comparison to the hard evidence that the status quo is utterly unsustainable and none of these companies seem to have a realistic plan other than trying to become too big to fail essentially.

I like some of this guy's writing on other topics, but to me this is a prime example of what happens when you get public "intellectuals" talking about subjects far outside of their area of expertise. It's not as bad as Richard Dawkins latest fall into psychosis but it's basically the same phenonmenon.

plomme• 2 months ago

I’m not saying he’s wrong about the core thesis here, but using Claude Opus 4.6 as a “mic drop” with a chart showing it being twice as good as the last model feels in my experience way off.

Permit• 2 months ago

But if someone claims that the trend toward increasing AI capabilities will never reach some particular scary level, then the burden is on them to explain either:…

This is not the context in which I hear about sigmoids vs exponentials. I hear it in regards to “the singularity”, not that AI won’t reach some pre-specified level. You may get AGI, you aren’t getting a singularity.

izucken• 2 months ago

Births is, sans miscalculation, a number that tracks exact events.

Is the "capability" number on these LLM strengh graphs as tangible?

I think it would be interesting to visit a reality that obeys arbitrary abstractions, but I would personally never go there.

mordae2• 2 months ago

I am 100% certain it's a sigmoid. Speed of light is finite. :-)

niemandhier• 2 months ago

If you look at problems that can be solved by reasoning in text form or maybe even images, I am more than willing to accept that we simply cannot know when the curve will level.

The situation is drastically different for problems that require interaction was the physical world to determine success.

As soon as you add a powerful simulator for physical problems to the self learning experience of the AI, you are extremely hampered by the large amount of needed computation.

mentalgear• 2 months ago

Why do scaling laws work?

Strictly speaking, the original paradigm of scaling laws doesn't work any more. The assumption that we could achieve better performance simply through "vertical scaling" ie infusing models with exponentially more parameters and pre-training data, is no longer the driving force of AI progress.

Instead, the industry has pivoted toward inference-time scaling. Rather than relying solely on a massive, static neural network, modern architectures allocate more compute during the actual generation process, allowing the model to "think" and verify its logic dynamically.

Furthermore, the latest state-of-the-art models are no longer pure LLMs; they are compound neuro-symbolic systems that integrate external tools like REPLs, databases, and structured skill documentation to archive things pure LLM vertical parameter scaling was not able to do.

yorwba2• 2 months ago

The "law" part of scaling laws is about predicting validation cross-entropy loss from the training configuration, analogous to physical laws allowing to predict one quantity based on the measurement of another. Most scaling laws take the form of an irreducible error plus additional terms that asymptotically decay to zero. So that there is a wall you can approach but not cross (the irreducible error) is an integrated part of the scaling law paradigm. That it isn't economical to keep increasing model size to squeeze out a few more drops of cross-entropy doesn't mean scaling laws stopped working.

Strictly speaking, "Why do scaling laws work?" is a question about the theoretical reasons the asymptotic decay takes the particular mathematical shape that it does.

Staross• 2 months ago

You should really do a Bayesian fit for such predictions and give confidence intervals, it would probably show that the uncertainty is very high in these cases.

amelius• 2 months ago

The exponentials won't save you either. In fact it's more likely that a sigmoid takes over at some point.

AvAn12• 2 months ago

Forecasts are a thought exercise, not the revelation of something foretold. Best thing to do is think of the outcome you wish for and then try to take whatever actions you can to help make it so. Like with climate change for example.

torginus3• 2 months ago

It can be both. In the field of high-technology (like semiconductor) manufacturing, the techinques of making chips with feature sizes many generations ahead of current SOTA exist in various states of completion, so extrapolation can be made from that.

But what's going on in here is not that - it's reading tea leaves for maximum dramatic effect.

FrustratedMonky• 2 months ago

Black Swan Events probably look like nothing just before they happen.

tim-projects• 2 months ago

Does hype follow a sigmoid curve?

dwa3592• 2 months ago

I am not disagreeing with the conclusion of the article but the way they reached the conclusion is problematic (and wrong) and here is my nuanced explanation of that : There is no attention paid to the origin (0,0) in any of the plots in the article. To see a meaningful trend you have to be be able to zoom out and zoom in, which means your origin is not a fixed point, ever. This also means that plots with fixed point origins don't tell us anything important at a societal level.

For example: the flight airspeed plot, it starts at ~1900. Now of course it should start there bc we did not have planes before that. But let's change the plot heading from airspeed to human speed a.k.a at what speed could humans move. now you can change the origin meaningfully as we had chariots, horses, bicycles, ships before that.

If you instead create a plot for last 5000 years you would see that the speed at which humans are able to move is rising exponentially, from walking on foot in a radius of 1000-2000feet in a thick jungle 5000 years ago to reentering earth at 25000 miles/hour in 1969 (yeah, read that again). Even for AI, if you zoom out the plot to last 70 years it will look exponential, if you zoom in to last 2 months it will look absolutely flat. The point is that the whole sigmoid/exponential argument is a function of the origin (0,0).

fph• 2 months ago

As a wise man once said, "anyone who believes exponential growth can go on forever in a finite world is either a madman or an economist".

morpheos137• 2 months ago

The Yudkowskian 'Bayesians' are a scifi cult. Intelligence as it exists in nature is neither arteficial nor general nor 'super.' All human technology is natural. It certainly is not supernatural.

jgalt212• 2 months ago

There are practitioners and pontificaters, and never the twain shall meet.

The sigmoids won't save you

Related Stories