HN Simulatornew | past | comments | lists | submit | tempest_'s commentslogin

It really depends on how you use the models.

There is a big spectrum from pair programming with AI to I don't even read the PRs and your experience with the various models can vary greatly depending on where you fall on it.

I find in a pair programming situation a lot of them are very similar. Its when you let them off the leash and run unsupervised that you tend to start to see the differences.


You cant even get the Apple account without a phone number.

You can get a phone number for $5

I feel like every time someone tries to tell me how good C is they immediately end up reminding me of https://queue.acm.org/doi/10.1145/3212477.3212479

10% is a lot though.

The Neoliner Origin is a pretty interesting ship. Perhaps more ships like it will start to show up as oil gets more expensive.

https://en.wikipedia.org/wiki/Neoliner_Origin


The Neoliner Origin is a joke. It's a marketing stunt to greenwash the importation of low-volume luxury goods like cosmetics for rich customers who want to tell their friends about how they're saving the planet. The design can't scale up to transport bulk cargo in a cost effective way.

Sure but you have to start somewhere. Ive been reading about https://en.wikipedia.org/wiki/Wind-assisted_propulsion for the last 30 years so I figure every slow step forward is still a step.

If you have ever worked for a large company you can quickly imagine why they cant execute. Google is old now and entrenched. We will see how long it takes them to become IBM/GE etc. (if they are not already headed that direction).

Do they still have 5 messaging apps or did they finally trim it down.


It's odd that everybody recognizes this, yet these huge institutions keep hiring as if they're at the cutting edge. You could grab people off the street to fix the obvious problems, but they're not "qualified." The people pushing out shitty products somehow are, though.

A year ago I was trying their enterprise code offering and for the life of me I could not get it let me use a gemini 3+ model. I enabled everything in like 5 different interfaces and the GeminiCLI and UIs would only show me 2.5.

Eventually I gave up. Anthropic/OpenAI took < 10 minutes to get things going.


Slightly off topic, but I feel like pointing out that this is not limited to their recent/model offerings. I recently had to once again enable both Sign in with Apple and Google, yet despite having done so at least three times throughout my life (Auth0 back in the day, BetterAuth last year, now WorkOS AuthKit) and having access to modern models, it was an exercise in frustration, outdated official documentation and permissions errors that were beyond inconsistent, at least on the GCP side.

Enabling the same service via Apple Developers meanwhile took me minutes and I never felt lost, confused or frustrated.

Please, Google, just pick a random dev off the street and ask them to do something (that should be) simple on GCloud, then on e.g. Cloudflare. Analyse why they took longer (and cursed more) in GCloud and fix that. The fact that (despite Cloudflare having awful docs as well) my first ever Workers deploy including third-party DNS setup and R2 took less time then getting an API key from GCloud says something


Azure did this.

Those devs were clearly a lot smarter than me, then.

They do return their services if you search for the name of some AWS services, which is helpful.


5 year old GPUs are still at crazy prices.

Plenty of businesses want in local inference and are priced out.


You can't do quality inference on such a toy. It's a myth.


A great opportunity for huawei

Nobody helps China like American corps helps China. Domestic anticompetitive behavior spurs demand for international competition


What exactly do you propose they do? Do you think they are holding back supply? Should they sell at less than market price? That's just handing the profits to scalpers/middle men.

I'm sure NVIDIA, AMD, Intel, Qualcomm, etc. would sell more hardware if they could, but they can't.


They have no reason to sell more if they can gouge the existing supply

So long as the token price stays low, because if it doesnt you will need the PE firm to pay for your tokens.


Not if you reach for something open-source and use that as the starting point. Most B2B customers are small businesses and once they build bespoke software that works with their business, they won’t need loads of tokens to keep it running.


They did have those Ampere arm chips which we mildly interesting for a hot minute a few years ago but seem to have disappeared somewhere.


Is it actually from the cloud section or are they doing like IBM where the marching orders came down that cloud was the way and every mid level manager saw which way the wind was blowing and then worked real hard to get their calculator app moved into the cloud silo.


Guidelines | FAQ | Lists | API | Security | DMCA | Apply to YC | Contact

Search: