firloop

I wasn't able to get this working, but I was able to use Google's mirror[0] just fine.

Just had to change

    FROM {image_name}

to

    FROM mirror.gcr.io/{image_name} 

Hope this helps!

[0]: https://cloud.google.com/artifact-registry/docs/pull-cached-...

++ had the same reaction as you. When I was picking up Swift, the interpreter was invaluable to check my understanding (although I nearly always needed to begin with `import Foundation` to have my code actually work).

It's still probably better to use Bun.

They've already forked Chromium. https://www.perplexity.ai/comet

I’m sorry for posting it.

Apple News+ link: https://apple.news/AlNSjMktVT9afQICM2yRVYw

What if they were offered $160mm and Tailscale countered with 4X the valuation, lowering the number of shares by 75%? Similarly, what if they wanted $40mm but the only deal on the table was $160mm due to ownership targets of funds that can actually write $40mm+ checks? It's hard to play these armchair games, even less so when the terms aren't known.

Any information on what search engine is powering it?

The post is about Deep Research which I don't believe is an API feature.

The post complains about token cost and invisible tokens, but the feature is only available to subscribers who pay a flat rate (not per token).

In the beta builds with Apple Intelligence, it is possible to turn the entire feature set off (including cloud calls). Don’t think you can keep the local features and skip the cloud though.

As I've mentioned on another thread, I just wish Waymo drove faster. They often don't keep up with the pace of traffic.

Writing this comment from Safari, which doesn't support third party cookies. The internet (and its ads) still work fine.

Yes, most human drivers go above the speed limit. I agree that it can be unsafe, but wouldn't a self-driving car be safer at going above the speed limit than humans? I feel like it should be OK for it to go 5-10 miles over, especially if that's what the flow of traffic is.

A rarely discussed negative to Waymo is that they drive slower than human drivers. Anecdotally they can be 10-25% slower than the rest of traffic, and it's not uncommon to see human drivers do unsafe moves to pass a Waymo.

Of course, but I would read it differently if the faked screenshots were wildly more absurd what was actually being generated (which in this case, they weren't!)

It's amusing to me that they spend so much energy explaining that there were a lot of faked screenshots... but in this very same post they admit the tool told people to put glue on pizza and eat rocks!

I like Nuun a lot

I use vast.ai somewhat often. It's great!

Yes, I had the same thought. Suspicious or coincidental.

The needed context seems to be that the person you are replying to is German. I don't love Schiphol either, but it beats any German airport (looking at Frankfurt in particular).

Depending on the specifics, 30% for a deep Apple partnership sounds like an amazing deal

You could use the OpenAI announcement: https://openai.com/blog/review-completed-altman-brockman-to-...

Some of these videos are not made with Sora. e.g. this is linked on the page: https://twitter.com/wxrdna/status/1758244945733865952 ... which was uploaded to YouTube over a year ago https://www.youtube.com/watch?v=f6YDKF0LVWw so we can safely assume it wasn't made with Sora. The tweet linked is probably a joke.

This wasn't always possible with TV; before DVR people would actually watch the commercials.

That’s the official API for HN: https://github.com/HackerNews/API

> This trick has simplified my project onboarding quite a bit. No more "You forgot to update your Email Address" requests from clients!

Does this happen in practice? Most commits on professional git repos that I work on are of the form <username>@users.noreply.github.com.

They're not charging for the model, they're charging for the UI.

I've submitted this post to this site before and I regret it. I think it's questionably true at most and is weird FUD.

Is this the Azure deployment of GPT-4 or something else?