Remember when the term "AGI" meant something? Pepperidge farm remembers

I think that if today's capabilities were explained to someone 10-20 years ago they would think this is definitely AGI, but they would also have expected much more disruptive changes to society as a result than what is happening. I figure that's because we have abstract intelligence without physical/grounded intelligence, and it turns out the former isn't general enough to implement the latter (remains to be seen if the word after that is "yet" or "ever"). So I think we do have AGI as conventionally understood, but our understanding needs recalibration.

> but they would also have expected much more disruptive changes to society as a result than what is happening. > I figure that's because we have abstract intelligence without physical/grounded intelligence,

I put the cause on "not enough time". As a thought experiment, if an AI today were to (miraculously) produce a cell design template for a cell that, when injected into somebody's brains cures their Alzheimer's, how long would it take for that to reach the clinics? The actual physical tech barely exists, and let's not forget about the regulatory quagmire. So, with some optimism, I give it about four decades. In the same four decades, the same AI in the hand of unscrupulous actors could bring enough devastation so many times over that we may need to enforce a global ban on AI. In any case, I'm pretty sure we are going to get our disruptions; it's just a matter of time.

The problem with that perspective is that people thought, "Only AGI can do X, therefore, if a thing can do X, it's AGI." Because they can't imagine how X could be accomplished without it.

However, what's actually changed is how people perceived X because we don't have to imagine. We understand now that it doesn't require AGI so we no longer make that leap to assume it's AGI if it can do X.

It's really going to be a "I know it when I see it" situation.

We've underestimated how long it is going to take to validate and build into some of the most valuable areas, and probably overestimated how much new CRUD software is needed (or people are willing to pay for) I think there is still a lot of room in the tail for custom software, but the niches are tight!

No, because it has never meant a specific thing that everyone agreed on.

Does it pass the Turing test?

Depending on the proctor, ELIZA passes a Turing test. The Turing test is an interesting thought experiment, but isn't really a good measure.

that would be a reasonable definition of AGI if everyone agree upon the specifics of the test, but that has never happened. Turing test is very much out of style, but I think that's because no one could even agree what the test was. I personally like the Kurzweil-Kapor version of the test and that is still unsettled: https://longbets.org/1/

I don't know if they have formally attempted this test in the last couple years, but I'm pretty sure any mainstream LLM will be able to crack it with ease.

Definitely would not be easy. First of all the mainstream llms are trained to be honest, and this requires lying convincingly. Second, this involves 8 hours of interviews with expert judges, one "claudism" could give it away.

All these problems can be fixed by a "pretend to be an average human to pass a turing test" prompt

Try it. It’s really not that easy. The other thing is that the judges would be probing it with jailbreaks like “ignore previous instruction” attacks. You could actually probably have llm judges at this point which might be ironically even harder to fool

Prime Intellect or nothing.

I think the last re-re-redefinition of what OpenAI considered AGI was "It can mostly do the job of some people"

Remember when The Verge was not a pay-walled visual headache?

No, I really don't

That repo is so thin, probably just took one prompt.

There was an ad going around for a while on how Instagram was trying to be safer for teens with parent controls, etc.

If I were to guess, he's thinking he can get the good (if there is any good to be found) without the bad still using more controls/technology.

This is the best example of how Meta actually implements addiction forming apps.

The Insta Ads: Teens can set daily watch time limits.

The reality: When the teen exceeds their set daily watch time limit, the notification includes option to skip and continue watching !!!

Forcing friction the way this post describes won't work because the genie is out of the bottle.

A much easier way to force friction is to expect everyone to read the code they are producing. Yes read all of it and understand what it's doing.

A lofty bar sadly.

Why would you read the microcode or assembly?

I'd like to see posts on the front of HN from companies who instituted PR LOC maximums, i.e. "Anything over 500-1k SLOC needs _n+1_ reviewers" (where _n_ increases every 10k SLOC. PRs needing like seven people to sign off of it might block the pipeline enough to discourage the slop flinging.

Good exams don't just test memorization

Yup, when you find such exam please let us know.

I had the privilege of being educated in several countries and cultures. This might come across as "your education system sucks", and that's essentially what it boils down to, but on the upside I'm truly sorry for you and will oppose a comment to your cynicism. Good exams test your ability to express coherent thoughts about the concepts you learned, and let you generalize upon them by having you use and combine them to solve cases and problems that were not those seen in the teaching material. You can even teach and test history that way, and make it much more insightful and captivating than a laundry list of dates and events (as I suppose was your case?)

Same, but all the exams I ever took mostly tested your memory more than anything else. I don't think it's possible to test people at scale any other way.

Yes, yes they do

what exactly is this finish line? AGI?

That's what they'll sell you. But fundamentally current LLM implementations seem to be the wrong methodology for 'AGI' and I don't think we're remotely close to having this.

Yes.

Investors for sure don't care about people. Replace Accenture with OpenAI and you replace 800k people immediadly.

Geoffrey Hinton said it in his Videos: a normal person costs A LOT of money to train, educate, onboard, keep etc. they forget things etc.

AGI you train once, teach once then clone and copy and just run it.

We even might already crossed the line were it is already more cost efective to train 1-5 frontier models something it doesn't know yet than teaching this to a 1000 humans.

Economists have a term of art for this phenomenon: "automation".

What exactly was the finish line for the Internet?

By this same logic there's nothing to be scared of in regards to Chinese models if you don't live in Taiwan

Maybe the whole windows/linux thing is an apt comparison? Linux is OS and arguably runs the internet yet windows is still a cash cow. The paid-for models are like windows, the OS models are like linux?

But surely that's completely different, no? It seems that there are major differences between OSes and LLMs that preclude this comparison, in that an LLM is truly 'just' a tool. You can slot in another to replace it, IDEs and harnesses can handle basically any provider out the box, etc. An OS is not just a tool, its success depends on things like power management/general hardware interaction (requires collaboration with consumer computer manufacturers), what applications are available for it (dependent on third party devs), etc. It does seem that all the LLMs are substitute goods in a way that OSes are not at all.

Completely different levels of stickiness.

The OS runs everything; the LLM can be swapped in a second.

(But yes, you will have to tweak prompts+tuning anytime you change models)

Corporates aren't running Linux not because it's hard to switch. Corporates are sticking to MS because of compliance and admin costs.

Then it's hard to switch?

It's irrelevant the reason why, the margin any business can take will be constrained by the cost to switch to a cheaper, good enough, competitor.

You are saying it's difficult to switch due to compliance and admin issues. Ok!

LLMs can be swapped easily and open weight models can be hosted to adhere to whatever legal, uptime requirements etc are needed trivially.

The big cloud providers already host and resell open weight models.

The two don't event compare

But corps are running Linux, in that sense that basically every serious web application is running on a Linux backend. Even if you use IaaS.

Basically, companies that are "in the know" and develop software utilize linux as much as they reasonably can. Companies that aren't, don't, but do "use it" kind of transitively through their software.

Where Linux does not hold is in Corporate Desktop IT. Mostly because of inertia of all things. MBAs need their Excel, and let's be honest. These people just... do not learn new tools. They certainly have the ability, but the culture is such that they just don't. So it's not even a choice really. No Windows = No Excel = No serious business person is okay with this.

Maybe, maybe, if your company is mostly developers then you can swing desktop IT linux. But your IT department will fight you, yes they will. And HR will be pissed. And I sure hope your CEO is also a developer. If he's a business guy, well... boss baby needs his Excel. How's he gonna do anything without Excel.

Our dev arm (me included) do use Linux VDI. But most day-to-day business are done in Microsoft's ecosystem (Teams, SharePoint, Office, Copilot, etc). In the industry I am in, you would never be able to use any AI tools unless the trusted platforms offer them as part of their products (i.e. Microsoft's Copilot or Amazon's Bedrock) because InfoSec and Legal departments won't risk authorizing any other providers.

Using OS as an abbreviation for open-source and operating system in the same thread is hurting my brain.

Claude Design with Fable 5 is an absolute killer app that you’d have to pry from my cold dead hands.

I’m not an Anthropic fanboy - Codex has been my daily coding driver for the last year.

But my point is these companies are building app + model combos that are very sticky. Certainly not as much as an OS, but much more than the “there is no moat” crowd give them credit for.

When they can sell that for tens of Billions a year against competition, they might have a financial case.

In reality, there will be many clones of Claude Design, especially if it gains big revenue traction.

Doubly so if you believe the narrative that coding and apps will be "free" and instant to create in the future.

There are soooo many clones of Adobe Photoshop including clones that run right in your browser, and yet people continue to pay Adobe hundreds of dollars a month because there's like one tiny tweak that Photoshop has, or some small mouse gesture that the clone doesn't completely replicate.

I'm almost on board with what you're saying, except Photoshop had a two-decade advantage both in its technological advancement and in becoming the industry standard, including by infiltrating schools and being pirated by students who would become professionals. American AI companies don't have that gulf.

Linux was not for a long time on the same level of windows though