Remix.run Logo
MangoCoffee a day ago

>Musk admitted to distilling their models.

what's the moat for AI labs like OpenAI and Antrhopic as they seek trillion IPOs. a lot of people said Chinese models also distill US LLM models. if it is this easy to do. how do US AI labs justify asking for trillion?

consumer451 a day ago | parent | next [-]

Well, they will stop releasing their latest models, and just attempt to eat all software themselves.

Otherwise, it will be competitors distilling, and the government getting upset.

It's hard for me to see a different future.

And, this really sucks.

If this thesis is true, I wonder what the final YC batch number might be.

lygaret a day ago | parent [-]

I believe if this was true, we'd already be seeing vibed stuff succeeding everywhere, and pricing out incumbents, but I guess I'm not seeing this? Like at all?

Even if they have Astra or whatever completely to themselves, how do they deal with the -product- side? Or sales, or account management?

They've been focused on juicing model's coding capabilities, but it's absolutely -not- "gen ai" enough to be doing the whole thing in agents, even 3 or 5 years from now, if only because there's so much context we _can't_ give to these models, without reverse centauring ourselves with cameras and mics and oodles of compute everywhere, and so far that hasn't exactly been playing out the way the frontier labs and the singularity folk hoped it would (meta glasses? humane? really?)

fma a day ago | parent | next [-]

>we'd already be seeing vibed stuff succeeding everywhere, and pricing out incumbents, but I guess I'm not seeing this? Like at all?

As a small business owner I see this. Random people contacting me to sell a software where I can instantly tell it is vibe coded. Subscription prices half or 1/4th of incumbents. Domain names registered in the last few months.

dgellow 19 hours ago | parent [-]

Yeah, but do they have customers?

halcy a day ago | parent | prev | next [-]

They don’t need to make you or me, people who actually make software, believe that every anthropic employee is actually and in fact, right now, a „10x software factory controlling thousands of agents“, they need to make investors and CEOs believe, and then hope that making that true is possible and that they can keep the fiction up long enough and get enough money and infra built out to make it actually true.

oblio a day ago | parent [-]

Like Uber replaced all drivers with self driving cars [1] in its 20 years of operation? Just because I want to paint the moon pink, doesn't mean it's feasible.

[1] If anything, Waymo has a higher chance and global Waymo adoption is probably 20 years into the future, at least.

reaperducer a day ago | parent [-]

Like Uber replaced all drivers with self driving cars [1] in its 20 years of operation?

They're trying. I see Uber robotaxis almost daily.†

I haven't bothered to see if they're still in training or actually taking passengers.

https://lucidmotors.com/stories/lucid-nuro-uber-partner

oblio 2 hours ago | parent [-]

I know they're trying. Trying is not the same as succeeding, which was my point.

citrin_ru a day ago | parent | prev | next [-]

Just guessing - companies which not outsource development to likes of infosys will outsource to OpenAI instead.

biql 18 hours ago | parent | prev | next [-]

One thing I see is design often getting worse, tasteless, how small but important details are just not thought out.

ben_w a day ago | parent | prev [-]

> I believe if this was true, we'd already be seeing vibed stuff succeeding everywhere, and pricing out incumbents, but I guess I'm not seeing this? Like at all?

I'm substituting my own[0] vibe coding for buying[1] apps. Language mini-games to help with German? A few prompts. Fluid dynamics simulation for an airzooker? Vibed. A web app listening for a MIDI keyboard, upon which you can drop some .midi files, and get a rhythm action game to learn the piano? Vibed. Webcam for my Raspberry Pi? Vibed. Getting Marathon 2 (well, the open sourced and upgraded engine, Aleph One) working as a web app? Vibed. Isochrone maps? Vibed.

Half of this I can even get done with the free models.

> without reverse centauring ourselves with cameras and mics and oodles of compute everywhere, and so far that hasn't exactly been playing out the way the frontier labs and the singularity folk hoped it would (meta glasses? humane? really?)

Yeah, so fortunate that cameras are expensive and there aren't 6 on my table right now between laptops and phones. :P

Seriously though, what's saving humanity collectively from everything getting automated from the panopticon we'd already built before Transformer models got good enough for even the most basic of classification and translation tasks, let alone anything we now use them for, is that machine learning takes an obscene number of examples before getting competent. Any living creature that needed so many examples would starve to death before learning how to eat.

This difficulty is why, for all the billions of miles that Tesla cars have collectively driven, perhaps pushing trillions now, they're still not sold to the public without steering wheels. Tesla claim to make such vehicles now in the form of the Cybercab, but they're not for sale, and even then some of the pictures that get in the press still show steering wheels.

[0] if you can call anything vibe-coded "my own"

[1] or worse, given the popularity of subscription models in this era, leasing some SaaS

jurgenburgen a day ago | parent [-]

> I'm substituting my own[0] vibe coding for buying[1] apps. Language mini-games to help with German? A few prompts. Fluid dynamics simulation for an airzooker? Vibed. A web app listening for a MIDI keyboard, upon which you can drop some .midi files, and get a rhythm action game to learn the piano? Vibed. Webcam for my Raspberry Pi? Vibed. Getting Marathon 2 (well, the open sourced and upgraded engine, Aleph One) working as a web app? Vibed. Isochrone maps? Vibed.

When do you have time for all of this? I don’t mean the vibing part but the using the app part.

I agree that agents are super good for one shotting throwaway code for tasks that would have required manual human actions previously but I would never have bothered buying an app for that.

Not having to deal with IT support for relatives is a win though since now I can just throw it at an LLM!

ben_w a day ago | parent [-]

> When do you have time for all of this? I don’t mean the vibing part but the using the app part.

Mix of this being spread over more than a year, that I'm not doomscrolling because HackerNews and Telegram are my main social media presences, and being unemployed/prematurely retired (which one depends on what one thinks of €1k/month passive income and no rent).

dominotw a day ago | parent [-]

Thats amazing. Family and kids path is getting less and less attractive everday. hobbies > family.

ben_w 21 hours ago | parent [-]

FWIW, I regret not having had kids yet.

I am also weirdly unmotivated by opportunities to spend money, which is both how I got this passive income and lack of rent, and why it's borderline enough for me. FIRE is very easy when all your working life, you only spend rent+50%, the rest of your paycheque going to savings and investments; but most people can't do this.

afiori a day ago | parent | prev | next [-]

Markets and rationality are sometimes just acquaintances

hparadiz a day ago | parent [-]

Oracle makes billions selling SQL when you can just use a free version. This is no different. Enterprise has enterprise needs. It's really not that complicated or irrational.

afiori a day ago | parent [-]

Sure but each of them is being priced as if each one is gonna have 90% marketshare in the future.

WD-42 a day ago | parent | prev | next [-]

Anthropic wants you to think that models that distill from them would be worthless otherwise. It’s not true, it’s just one part of the process. The whole narrative that Chinese and other models are only good because they distill is nonsense.

hirako2000 a day ago | parent | prev | next [-]

Distilling merely saves an expensive part of the process: hundreds of millions. So even if distilling wasn't a thing, what justifies trillions ?

dgellow 19 hours ago | parent | prev [-]

There is no moat, other than the branding