polskibus

It could be that some researchers can get access to latest models or no limit accounts only if they approve of usage of their data. It would be interesting to know which researchers have had which settings/setups.

Are you saying researches don’t choose the option to not use their data for training, or they do select it but OpenAI disregards it?

Can you provide a reference to that?

The his will likely put a strain on neighbouring countries energy supply who used to buy that surplus before. Prices will go up even more.

What compact camera would you recommend to get to take pictures that look noticeably better than iPhone’s? I mean without running around with a bag of lenses.

Minority Report anyone?

What should be done to keep it on the right level forever ?

MS SQL has been available on Linux for around 10 years now.

So is the magic basically to use ZFS and everything should work fine?

Would such approach work for MS SQL?

By „with each pass” do you mean within the same session, or with new session (context window) each time?

Look at the demography though. Poland will fall very hard, faster than its neighbours.

How big is the 996 crowd compared to the total tech employment in SF? Is it really that popular?

Can this technique help with fixing a bug on master then having to merge it easily to older release-branches , without resorting to cherry-picking (and losing all history)?

Why is the old image so much more blue? Did pollution increase cause this change in color over time?

Don’t forget WhatsApp. Kids are allowed to have WhatsApp as messaging but they get fed videos there too. There is no way to really disable them . Also this be allowed as parental supervision, not something that kids can override.

Is it possible to run minikube inside ? I’d love to use it for ephemeral clusters for testing .

On the other hand, shouldn’t there be a policy forbidding use of HN data for LLM training? I would certainly be more encouraged to participate, if I knew that the content I provide for free is not used to train LLM that is later sold by a company valued hundreds of billions. Perhaps there are others who feel the same.

Surely MinIO dual-licenses its software so paying customers get commercial license?

So did the Linux compiled with this compiler worked? Does it work the same as GCC-compiled Linux (but slower due to generating non optimized code?)

Does this feed DuckDb continuously data from transactional workloads, akin to what SAP hana does? If so that would be huge - people spend lots of time trying to stitch transactional data to warehouses using Kafka/debezium.

BTW, Would be great to hear apavlo’s opinion on this.

What's the SOTA for on-prem Postgres, in terms of point-in-time-recovery? are there any well-tested tools for it?

Lean4 seems to be very popular in the Math Olympics-solving AI startups (Harmonic etc).

Wouldn’t it make more sense to write the same functionality using a more performant, no-gc language? Aren’t competitors praised for their CLIs being faster for that reason?

One of the authors, Adrian, is a very interesting person. Got his PhD at 21, started CS studies at an age when his peers were starting high school. Knowing some of his previous achievements, I’d say his work deserves at least some curiosity.

I posted it 19 days ago. It it didn’t get any traction, I wonder why. https://news.ycombinator.com/item?id=45453119

Can this be used in a public cloud provider , to speed up VM provisioning in CI/CD pipelines? Im looking for ways to speed up app provisioning for e2e tests.

I think the post, while extensive, missed one important issue.

The fact then when we read others' code, we don't remember/integrate it into our thinking as well as we do when we're the authors. So mentoring "AI Juniors" provides less growth then doing the job, esp. if it is mostly corrective actions.

Was hoping for a more data-driven diagnosis. The reality is much different, smaller orgs can move faster than large. It is definitely not possible in areas that require huge CAPEX or OPEX like AI, but in many other areas it happens often.

Reminds be of cyberpunk dystopia. Several corporations like Arasaka and Militech defacto ruling the world.

By drag I mean 3-4 hrs until all e2e passed on an ephemeral k8s stack on a vm.