squeaky-clean

You're still hung up on this idea that violating boundaries and being a man are synonymous, and that what makes it creepy is how attractive you are.

Whether someone is attractive or not, they're perfectly capable of not bothering someone who doesn't want to be bothered.

You're basically saying "if an unattractive man who doesn't respect boundaries is creepy, but also an attractive man that doesn't respect boundaries is creepy, then that means all men are creepy."

It's very telling that you think creepy and ugly are you synonyms and non-creepy and attractive are synonyms.

Those guys are creepy too.

If they're a non-creepy guy, it's because they don't behave like a creep. Not because they're handsome.

> FFS, git is a tool for programmers.

We already said it has a garbage UI, you don't need to repeat it

But jokes aside, usually the reason something is considered a "tool for programmers" and not just a tool is because the UI sucks.

If you look at any other details of this it is very clearly vibecoded.

> Long-running tasks like batch mode completions and agent sessions may incur overages beyond your project spend cap.

> Billing data processing times can be delayed in AI Studio, up to around 10 minutes. You may experience overages beyond your project cap if billing data hasn't processed before more charges are accrued.

https://ai.google.dev/gemini-api/docs/billing#project-spend-...

That's a soft cap, not a hard cap

I'm reading a book by Asimov at the moment, printed 1989. The first page is an ad for a Robert Silverberg book adaptation of the Nightfall short story. Pages 2-3 are ads for other Asimov books. The last 5 pages are an excerpt of Edge of Eternity. And the final page is an advertisement for Voice of the Planet, a "5 episode TBS miniseries starring William Shatner and Faye Dunaway"

"My AI writing isn't slop. It's just as good as buzzfeed lists or celebrity gossip articles."

Those are just awful compared to the side view of a pelican on a bike.

Which of these programs is easier to review

  {x{x,sum -2#x}/0 1}

or

  def f(n):
      if n <= 1:
          return n
      else:
          return f(n-1) + f(n-2)

They're both the same program

There's already feedback in the gitlab issue that the top commenter linked. Their HN comment isn't really adding anything

Hit me with the downvotes, but the only thing their comment has contributed to is causing arguments.

Just today I asked Claude Code to generate migrations for a change, and instead of running the createMigration script it generated the file itself, including the header that says

  // This file was generated with 'npm run createMigrations' do not edit it

When I asked why it tried doing that instead of calling the createMigrations script, it told me it was faster to do it this way. When I asked you why it wrote the header saying it was auto-generated with a script, it told me because all the other files in the migrations folder start with that header.

Opus 4.7 xhigh by the way

If they're shuttering this farm because peaches are no longer profitable, why would someone else pay for these trees?

I like Codea for iOS, though the free version has a soft-limit at 500 lines. If a project gets bigger than 500 lines you can still run code but it'll nag you to upgrade.

That doesn't get you any Claude usage. Claude models obviously aren't open, but equivalent models to Opus take about 400GB of memory to run.

You can run the versions with fewer parameters or quantized weights but depending on how much quality you're sacrificing, now you'd have to compare the price against cheaper Claude models like Sonnet.

I've had Opus 4.6 1M and Sonnet 4.6 1M for months now on Bedrock.

Basically any cloud connection to Amazon will be closed on these devices. The only way to add books will be USB.

Ctrl+F. Highlighting, notes, and bookmarks with the ability to hide them. Reading in a dark room. Adjustable font size. Slim size (that's related to portability but I can also read an 800 page book in my bed without getting my arms tired).

I have a Kindle that took a fall about 8 years ago and the wifi has never worked since then. I've been able to load books onto it using USB with no issues.

This comment is tangential to their point that a transformer architecture can or cannot be functionally equivalent to a human brain. Practicality of those limitations is a different discussion

This is false. You can ask it to spell out strawberry and count the letters and it will still say 2 (it's unable to actually count the letters by the way). The only way to get a model that believes strawberry has 2 R's to consistently give the correct answer is to ask it to code the problem and return the output.

In fact, asking a model not to repeat the same mistake makes it more likely to commit that mistake again, because it's in it's context.

I think anyone who uses LLMs a lot will tell your that your steps 3 and 4 are fictional.

Are you aware that a lot of model tool calls are useless and a smarter model could avoid those?

Are you aware that output tokens are priced 5x higher than input tokens?

If you and I both pay $10/mo to listen to Spotify, and we are the only subscribers. If I listen to 1 song by Sabrina Carpenter, and you listen to 99 songs by Taylor Swift. Then of our $20 (after Spotify's share) 1% of the money will go to Sabrina and 99% of the money will go to Taylor. Because Taylor was played 99x more than Sabrina. Even though for both of us as users, our respective artist was 100% of our listening.

It doesn't calculate your amount of listening and determine the payout based on that. All listens are pooled together and all subscription money is pooled together. And the payout is determined based on that.

Those are for devices not in lockdown mode.

They are saying web based solutions often out perform LÖVE, even though you would expect the opposite because LÖVE doesn't have the bloat of a browser engine.

And it's worse than spam when someone is posting incorrect things and people are downvoting people questioning it. As another user has already posted, the Iron Dome does not use the same radar they are talking about and is not "blind"

Settled out of court does not mean the lawsuit never went to court. It means the settlement happened outside of court. Every lawsuit has to go to court, that's how you file a lawsuit. If it isn't sent to a court it's just words in a document.

SQLite had 6 CVE's last year. FreeType had an RCE bug published last year. libcurl's last CVE reported was 2 weeks ago. Libpng had 4 vulnerabilities published this year....

The comment you're replying to literally has an example of an internet calling service being fined $20,000 for not properly directing 911 calls.

I guess Vonage should try to appeal the case and say pocksuppet said they're not required to do that.

Now no one has a billion dollars but they have 100 companies they control each worth 10 million.

These lawsuits have to be designed with the idea that the people with the most resources will try to exploit them, and the people with the least resources will be unable to.

10 years ago the most popular 100 videos on YouTube were all pop music videos. Justin Bieber had 3 of the top 10.

The youtube algorithm has been personalized for much more than 10 years and has never prioritized any kind of lectures or artful films over anything else it thinks a viewer will watch. You're asking for them to bring back an era that never existed.

If you're not getting those sorts of recommendations it's because you ddon't actually watch that kind of content, or you're removing your history.