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

OpenAI declares 'code red' as Google catches up in AI race

theverge.com

Post reply on HN
Print view View original

791–800 of 960 posts

Re: OpenAI declares 'code red' as Google catches up in AI race

#791

Post by chasd00 »

kshacker wrote: post #446
bovermyer wrote: post #440

Earlier quoted context omitted.

I have never considered trying to apply Claude/Gemini/etc. to Fortran or COBOL. That would be interesting.

I was just giving my history :) but yes I am sure this could actually get us out of the COBOL lock-in which requires 70 years old programmers to continue working. The last article I could find on this is from 2020 though: https://www.cnbc.com/2020/04/06/new-jersey-seeks-cobol-progr...

Or you could just learn cobol. Using an LLM with a language you don’t know is pretty risky. How do you spot the subtle but fatal mistakes they make?

Re: OpenAI declares 'code red' as Google catches up in AI race

#793

Post by randomNumber7 »

binkHN wrote: post #568
alecco wrote: post #317

Earlier quoted context omitted.

SemiAnalysis said it last week and AFAIK it wasn't denied. https://newsletter.semianalysis.com/p/tpuv7-google-takes-a-s...

This is a really great breakdown. With TPUs seemingly more efficient and costing less overall, how does this play for Nvidia? What's to stop them from entering the TPU race with their $5 trillion valuation?

If you look at the history how GPUs evolved:

1. there had be fixed function hardware for certain graphics stages

2. Programmable massively parallel hardware took over. Nvidia was at the forefront of this.

TPUs seem to me similar to fixed function hardware. For Nvidia it's a step backwards and even though they go into this direction recently I can't see them go all the way.

Otherwise you don't need cuda, but hardware guy's that write verilog or vhdl. They don't have that much of an edge there.

Re: OpenAI declares 'code red' as Google catches up in AI race

#794

Post by steveBK123 »

atonse wrote: post #546
steveBK123 wrote: post #314

Earlier quoted context omitted.

If the subscription cost 5x as much would you still pay and feel you are getting such a great value?

I pay $100/month for Claude Max, and I've already said it, I would go up to $500 a month and wouldn't hesitate for a second. I'd probably start to hesitate for $1,000 maybe, only cuz I know I wouldn't be able to use it enough to maximize that value. But I might still suck it up and pay for it (I don't use it enough yet to need the $200/month but if I started hitting limits faster, I would upgrade), or at that point s…

Right I am sure some find it is worth 5-10x the cost.

The challenge is that if the numbers are accurate they need 5-10x to break even on inference compute costs, before getting into training costs and all the other actual overhead of running a company like compensation.

Will everyone be willing to pay 5-10x? Probably no.

Will half of users pay 10-20x? Or a quarter pay 20x++?

Or we end up with ads … which already seem to be in motion

Re: OpenAI declares 'code red' as Google catches up in AI race

#795

Post by hansmayer »

bgirard wrote: post #290
stack_framer wrote: post #247

Earlier quoted context omitted.

I can generate $20 billion in ARR this year too! I just need you to give me $100 billion and allow me to sell each of your dollars for 0.2 dollars.

It's a fun trope to repeat but that's not what OpenAI is doing. I get a ton of value from ChatGPT and Codex from my subscription. As long as the inference is not done at a lost this analogy doesn't hold. They're not paying me to use it. They are generating output that is very valuable to me. Much more than my subscription cost. I've been able to help setup cross app automation for my partner's business, remodel my ho…

> It's a fun trope to repeat but that's not what OpenAI is doing.

This is literally what OpenAI is doing. They are bleeding cash, i.e. spending more than they earn. How useful it is to you is not relevant in the context of the sustainability. You know what is also super useful to some people? Private yachts and jets. It does not mean they are good for the society as a whole. But even leaving out the hollistic view for a moment - their business model is not sustainable unless they manage to convince the politics to declare them national infrastructure or something like that, and have taxpayers continue to finance them, which is what they already probed for in the last months. Out of interest, why would you want ChatGPT plan your trip to Japan? Isn't planning it yourself a part of the excitement?

Re: OpenAI declares 'code red' as Google catches up in AI race

#796

Post by hansmayer »

usef- wrote: post #572
bloppe wrote: post #318

Earlier quoted context omitted.

To be fair, I would get a ton of value out of someone selling dollars for 20 cents apiece. But ya, OAI is clearly making a ton of revenue. That doesn't mean it's a good business, though. Giving them a 20 year horizon, shareholders will be very upset unless the firm can deliver about a trillion in profit , not revenue, to justify the 100B (so far) in investment, and that would barely beat the long term s&p 500 average…

People did say the same thing about Youtube, which was unprofitable and extremely expensive to run in the early years. I remember thinking everyone would leave once ads were added. At youtube's ad income rate (~$13/year), the current (but growing) ~800 million chatgpt users would add ~$10 billion. At facebook's rate (~$40-50/year) $32-40 billion. Potentially, an assistant would be more integrated into your life than…

Minor difference : YT does not cost literally a human trip to Mars and back to operate

Re: OpenAI declares 'code red' as Google catches up in AI race

#797

Post by qdotme »

matwood wrote: post #708
sigmoid10 wrote: post #688

Earlier quoted context omitted.

Making the hardware is actually the easy part. Everyone and their uncle who had some cash have tried by now: Microsoft, Meta, Tesla, Huawei, Amazon, Intel - the list goes on and on. But Nvidia is not a chip company. Huang himself said they are mostly a software company. And that is how they were able to build a gigantic moat. Because noone else has even come close on the software side. Google is the only one who has…

Are you saying that Google, Meta, Amazon, etc... can't do software? It's the bread and butter of these companies. The CUDA moat is important to hold off the likes of AMD, but hardware like TPUs for internal use or other big software makers is not a big hurdle. Of course Huang will lean on the software being key because he sees the hardware competition catching up.

Essentially, yes, they haven’t done deep software. Netflix probably comes closest amongst FAANG.

Google, Meta, Amazon do “shallow and broad” software. They are quite fast at capturing new markets swiftly, they frequently repackage OpenSource core and add the large amount of business logic to make it work, but essentially follow the market cycles - they hire and layoff on a few year cycle, and the people who work there typically also will jump around industries due to both transferable skills and relatively competitive competitors.

NVDA is roughly in the same bucket as HFT vendors. They retain talent on a 5-10y timescales. They build software stacks that range from complex kernel drivers and hardware simulators all the way to optimizing compilers and acceleration libraries.

This means they can build more integrated, more optimal and more coherent solutions. Just like Tesla can build a more integrated vehicle than Ford.

Re: OpenAI declares 'code red' as Google catches up in AI race

#798

Post by Dumblydorr »

chem83 wrote: post #413
twothreeone wrote: post #174

The way I've experienced "Code Red" is mostly as a euphemism for "on-going company-wide lack of focus" and a band-aid for mid-level management having absolutely no clue how to meaningfully make progress, upper management panicking, and ultimately putting engineers and ICs on the spot to bear the brunt of that organizational mess. Interestingly enough, apart from Google, I've never seen an organization take the actual…

> fire mid-management and PMs to prevent the same thing from happening again Firing PMs and mid-management would not prevent any of code reds you may have read about from Google or OAI lately. This is a very naive perspective of how decision making is done at the scale of those two companies. I'm sorry you had bad experiences working with people in those positions and I wish you have the opportunity to collab with gr…

Yeah the reflexive anti-PM anti-management stance posted above is typical here and of devs in general.

In theory, some engineers think they are perfectly capable of doing all the PMs work and all their own.

If they’ve never worked with a truly good PM, that’s a shame, they’d likely get more work done in a more timely fashion. I’ve worked with around 10 different PMs, the best kept stuff on track and aided with collaboration, reqs management, soft skills, handling tough customers, etc. they free up devs to do more dev work and less other work.

Re: OpenAI declares 'code red' as Google catches up in AI race

#799

Post by captainbland »

binkHN wrote: post #568
alecco wrote: post #317

Earlier quoted context omitted.

SemiAnalysis said it last week and AFAIK it wasn't denied. https://newsletter.semianalysis.com/p/tpuv7-google-takes-a-s...

This is a really great breakdown. With TPUs seemingly more efficient and costing less overall, how does this play for Nvidia? What's to stop them from entering the TPU race with their $5 trillion valuation?

Nvidia is already in the TPU race aren't they? This is exactly what the tensor cores on their current products are supposed to do, but they're just more heterogeneous GPU based architectures and exist with CUDA cores etc. on the same die. I think it should be within their capability to make a device which devotes an even higher ratio of transistors to tensor processing.

Re: OpenAI declares 'code red' as Google catches up in AI race

#800

Post by tomrod »

fzzzy wrote: post #770
tomrod wrote: post #755

Earlier quoted context omitted.

Current Gen LLMs are not breaching the moat yet.

Yeah they are. llama.cpp has had good performance on cpu, amd, and apple metal for at least a year now.

Thw hardware is not the issue. It's the model architectures leading to cascading errors
Post reply on HN