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

Apple Foundation Models

platform.claude.com

Post reply on HN
Print view View original

211–220 of 244 posts

Re: Apple Foundation Models

#211

Post by tedggh »

joenot443 wrote: post #100
tedggh wrote: post #95

Earlier quoted context omitted.

I use both Claude and Codex and don’t see any meaningful difference between the two. My use case is modeling semi complex physical processes (energy and manufacturing) in code for simulations. I also have to do a good fair of automation via scripting in Python or PowerShell for manipulating data as well as legacy code analysis (C, Fortran, COBOL). Given I provide the models with the information and documentation they…

Did you find much of a difference between Fable and Opus?

I have used Fable only once to do an in depth codebase review of a complex system. I asked it to flag deviations from a particular design and also compile a list of vulnerabilities. It took about 15-20 minutes. The result was very similar to Codex for the most critical findings, different suggestions on how to address them but it found exactly the same critical issues as Codex. This is still not a good test to evaluate Fable. But my feeling is that the latest models are all pretty good and now it comes down to your personal setup and workflow, that’s where you can get the productivity gains IMO. It’s like picking between MacOS or Windows as development environment. For some Windows sucks and for a some is the opposite, but both groups of people can be equally productive if they know their environments well and know how to go around their respective limitations.

Re: Apple Foundation Models

#212

Post by dlev_pika »

harrouet wrote: post #49

This is Apple commoditizing LLMs while keeping control of the UX. They are a hardware company and will keep selling the best machine for AI use. Well done.

Apple’s play was a masterclass - unsure how deliberate it was, or how much of a choice thy actually had, but it’s turning out pretty well IMO.

Now if they can further reinforce their angle on Privacy, they might continue to be what they are (or more)

Re: Apple Foundation Models

#213

Post by ak_111 »

ksec wrote: post #94
alecco wrote: post #79

Earlier quoted context omitted.

In spite of their deeper pockets, massive datacenters, colosal amounts of user data, and hundreds of thousands of top developers, even Amazon, Meta, Microsoft, and Google are well behind. I think Evans is completely wrong. There are only 2 truly frontier models. (at least for now). And Anthropic seems to be leaving OpenAI behind so there might be only 1 in the near future. (which is scary/dangerous)

>I think Evans is completely wrong. I wish there was a case where I find Evans is wrong. As far as my memory served me, I failed to record a single one. I disagree that Amazon, Meta, Microsoft, and Google are " well " behind. If anything the frontier model advantage seems to be at best 6 - 9 months. And that the Chinese model are all doing well. One of Steve Jobs's line, "It is a feature, not a product." Even if Appl…

Just top of my head (and I don't even follow his takes that closely), just check his takes on Magic Leap which he consistently promoted using quite dramatic langauge (along with the entire AR space) and check how it panned out.

Re: Apple Foundation Models

#214

Post by tedggh »

axus wrote: post #166
tedggh wrote: post #72

Earlier quoted context omitted.

Benedict Evans may be right after all; frontier models look more and more like telecom companies in the 90s. Billions and billions of investment in infrastructure while others further up the stack captured all the value.

Last I checked the telcos made plenty of money in the 90s. Should Verizon be getting a cut of my Claude Pro subscription, since I use FIOS to access it?

I haven’t fact checked, but according to Evans big telecom builders didn’t make a lot of money after all the capacity investment. Some actually went bankrupt or got acquired as distressed assets. Big tech was very profitable monetizing that same infrastructure.

Re: Apple Foundation Models

#215

Post by nyrikki »

CuriouslyC wrote: post #144
yandie wrote: post #140

Earlier quoted context omitted.

I doubt that. What stops the Chinese labs from figuring it out? It’s not like these models are fundamentally different from each other

If all you have is the starting point and the finishing point, the lack of the path taken from one point to another limits your ability to train models that can efficiently recreate the work, and increases its cost enough that it's possible the US labs can progress capabilities faster than Chinese labs can distill that behavior.

[dead]

Re: Apple Foundation Models

#216

Post by johsole »

thewebguyd wrote: post #170
wolttam wrote: post #137

Earlier quoted context omitted.

I think it's highly likely that there will remain one or two companies on the very bleeding edge of AI development for the foreseeable future. But what I think a lot of people miss is that the market for the truly bleeding edge (developing bio-tech, building the most sophisticated software stacks (probably with a tilt towards simulation, GPU kernel optimization, etc)) is not the whole market. There's a plethora of us…

Anecdotal case in point, but writing mostly enterprise CRUD in C#, I've gotten plenty of mileage out of Sonnet, very rarely do I need to use Opus. Its somewhat of a myth that you need the most advanced, expensive model for software development.

There was a time when Opus was the only model really worth using, I think that was maybe 4.4 or 4.5, but I agree Sonnet is pretty good now and can be used quite often.

Re: Apple Foundation Models

#217

Post by awongh »

alecco wrote: post #79
tedggh wrote: post #72

Earlier quoted context omitted.

Benedict Evans may be right after all; frontier models look more and more like telecom companies in the 90s. Billions and billions of investment in infrastructure while others further up the stack captured all the value.

In spite of their deeper pockets, massive datacenters, colosal amounts of user data, and hundreds of thousands of top developers, even Amazon, Meta, Microsoft, and Google are well behind. I think Evans is completely wrong. There are only 2 truly frontier models. (at least for now). And Anthropic seems to be leaving OpenAI behind so there might be only 1 in the near future. (which is scary/dangerous)

That's true now, but long-term (maybe just a few years) it doesn't seem feasible for the status quo to continue from a financial point of view.

Spend for compute seems like it needs to increase to get the next iterations of models, and even if they IPO the money might run out before they can solidify their revenue streams.

All while Google just needs to survive long enough with their good-enough models and do it without really putting themselves in any existential financial risk.

And ideally the chinese models are also still there keeping everyone honest.

The true dystopic worst case is a Google monopoly on cutting edge AI.

Re: Apple Foundation Models

#218

Post by hylaride »

tedggh wrote: post #214
axus wrote: post #166

Earlier quoted context omitted.

Last I checked the telcos made plenty of money in the 90s. Should Verizon be getting a cut of my Claude Pro subscription, since I use FIOS to access it?

I haven’t fact checked, but according to Evans big telecom builders didn’t make a lot of money after all the capacity investment. Some actually went bankrupt or got acquired as distressed assets. Big tech was very profitable monetizing that same infrastructure.

Some went bankrupt, with Worldcom being the most famous example...though that was fraud. But even those that remained had large amounts of debt that never ends as there's always CAPEX for upgrades to networks to fund (both fixed and wireless). Now a lot of the debt is also from some of them going on media ownership adventures, but even those that didn't eventually got folded into larger companies (eg Sprint).

Most of the ones that survived did so due to being able to pick up distressed assets and at values that could then be profitably monetized - a move that it would not surprise me to see repeat itself in the LLM space (we'll see).

Re: Apple Foundation Models

#220

Post by 5701652400 »

WorldMaker wrote: post #172
swingboy wrote: post #77

Earlier quoted context omitted.

Does “the best machine for AI use” apply here considering these models are still server-side?

Apple's been trying to make the marketing appeal that "Private Compute Cloud" is also a hardware project. Given it seems to rely on low level details of device Hardware Security Modules, it's maybe even at least a little bit more than just "marketing spin".

looks like it is not "Private iCloud Compute" at all.

Anthropic literally says "Requests go directly from your app to the Claude API; Apple is not in the request path and does not see prompts or responses." — Apple straight up lied

Post reply on HN