lefty2

interesting story, but the fact that Apple and Nvidia were sueing each other doesn't seem any reason to drop all open standards. I would say they did it because of arrogance - ie. we're Apple and everything we do is better, therefore our graphics API is superior than any other

> they claim that two of those three quarters were profitable.

no they didn't. What they said was "positive adjusted operating income". The most important word there is "adjusted", ie. they are not profitable if you include all costs

Also, bear in mind the guy who wrote the article lives in the same flat as Sholto Douglas, an Anthropic researcher.

woosh

Interesting fact, you only could come third in the Olympics during the bronze age

Even GPT Images 2.0 was already the best image generator that I have tried, however it has one massive disadvantage: draconian censorship. Even rendering a gun in a scene will get flagged as "violent" and I need to create a battle scene... there's no way to do that with the censorship the way it is

The first version is already beating Rubin in the benchmarks, according to SemiAnalysis

So, Nvidia are backstopping $100B for the Ohio project, yet at the same time OpenAI are developing their Jalapeño processor, which will mean it will no longer need Nvidia GPUs for inference. That does not really make any sense to me.

> Even Deepseek at its absurd prices is a very healthy business. Regarding their return on capex multiple, their CEO said they make a six-fold profit on their compute capex with 10 month recuperation.

That's "theoretical profit" - in some imaginary world where the free subscribers would pay the top tier cost.

https://techcrunch.com/2025/03/01/deepseek-claims-theoretica...

Don't worry, if it escapes from the lab we can just blame it on the bats again

I understand now. Flagged means inappropiate for Hacker News. Yeah, they always flag any article that is anti-AI - even if it is well reasoned

flagged as what?

> I literally do not write code anymore.

but you are taking advantage of a subsidised service. What happens when developers are forced to pay the full price (i.e. $2000 a month instead of $200 a month)

Also the agreement can be can be terminated with only 90 days notice

yes, but not necessarily cheaper

The Chinese models are only temporarily cheaper, because they are subsidised the same way that frontier models are. Once those companies need to make money the subsidy will disappear.

According to the article they don't have to declare leases or GPUs until the datacenter becomes operational

I tried both unordered_flat_map and hopscotch map with the pathfinding algorithm that my game uses. Both were slower than regular unordered_map. unordered_flat_map about 33% slower and hopscotch was 390% slower

They would be jamming their own glide bombs and drones then. It would be more useful to jam Russian airspace to defeat Ukrainian drone attacks.

> If / when enough companies exit the market

That will only happen when the bubble bursts and those companies will exit by going bankrupt

Who is it written by anyway? The author's name is not given in the article

I wonder how many false positives there were. Typically this types of static analysis tools come up with a ton of potential bugs, but only a few of them are actual bugs.

Sora was losing 15M a day and it was running at least 3 months, so that's a total 1.3 Billion. That's a pretty expensive experiment. It sounds like a company with lots of VC to burn and no discipline. Even Jensen Huang accused them of lack of discipline in business approach.

also they need to pay back that in one year, so if OpenAI don't IPO this year they are screwed

> -1 still produces -1 instead of 0 That could be a problem depending how you are using it.

> The same trick can also be used for the other direction to save a division:

> NewValue = OldValue >> 3;

You need to be careful, because this doesn't work if the value is negative. A

> Yes, OpenAI is burning $8-12B in 2025

OpenAI actually burnt $12B in just one quarter (not for the entire year as he claims)

https://www.thurrott.com/a-i/openai-a-i/329108/openai-lost-1...

c++ 03 was a lot easier.

For instance, if you want to avoid unnecessary copy operations when returning a string, just return it in variable that you pass by reference (eg. void doSomething(string& str);) likewise avoid the vector class making unnecessary copies, simply by creating the objects on the heap and use a vector of pointers instead of values. It's a bit more ugly, but it works, and you don't need to read a 24 page blog to understand all the corner cases where it can go wrong. modern c++ is all about syntactic suger.

If it's untrustworthy, why do they prevent us from disabling "AI overview"?