janilowski

>“There’s not anything like this in the world,”

Must've never driven through the Alps I guess. There's something like this every couple km

You are correct; however, agents can and do find terrible bugs in some of the most popular software in the world made by some of the most competent engineers on the planet.

I assume what KingOfCoders suggested is to run an agent almost like an antivirus which I guess sounds somewhat interesting but you'd have to re-run it on every update and probably spend many more hours+tokens for such a system to work properly.

But perhaps once inference gets truly cheap repository maintainers should include AI security checks for new software as part of their pipeline? I think that would make sense. Especially for repos that are frequently pwned like npm!

Nice UI but untrusted email messages is something I'd rather open in [an uglier app with decades long reputation for being reasonably secure](https://www.thunderbird.net/en-US/)

By the way, if I was wanted, I'd just pull my shirt up over my face or wear a facemask

This is misleading in so many ways...

Their watchlist was only ~450 people, and it was restricted to people relevant to the transport police. The system was also overt and clearly marked, there were alternative routes that avoided the cameras and even then they were on 3-6 hours per day.

The headline suggests this system is not only invasive but also ineffective. While I agree it's a privacy nightmare, I also think those idiotic deployment conditions set them up for failure. At least we know there aren't many false positives, it could've been worse.

> It's sad to see F-Droid using the now-standard design convention of categorically refusing to draw lines between things.

Where exactly would the UI benefit from "drawing lines"? I'd say the visual hierarchy is rather obvious.

I'll say more: most separators people use in their UIs are completely unnecessary. They should be used sporadically when you don't have a better way of segmenting the interface.

> No visual differentiation between sections of the UI, no clear indication of what's tappable and what will happen on a tap

What section of the UI are you referring to exactly? From the screenshots in their blog post it seems to me like the visual hierarchy is preserved and without using the app at least I can form an immediate expectation as to the function of each button. (And what is clickable and what isn't)

> no indication of scrollable areas beyond content being cut off

What type of indication would you like to see here? I think most users intuitively understand that visual cue. An always-on scrollbar would introduce unnecessary clutter..

They’ve already made it obvious that the row is horizontally scrollable by leaving part of the next icon visible at the right edge. That's a clear (and standard) visual cue that there’s more content off-screen.

Perhaps we're in a transition period where we still mostly haven't figured out how to get AI to increase quality & productivity instead of just volumetric output.

That's my experience too. I've found OpenAI really quite generous with tokens. I sometimes wonder how some people manage to run out of them really. Do they just type prompts that much faster than me or use the highest reasoning mode for everything just because they can? Idk.

I generally agree with those reasons, although using a single thread may be less of an issue than it seems because of context compacting which should happen automatically when you're near the limit.

How do you manage to run out of tokens so quickly? I probably run more threads every working day, usually on medium, and I'm still below the 5x limits.

Do you use the official harness? OpenAI's models are generally best in class for token efficiency. It seems to me like they push for that much more than their competitors.

If I take a photo and then use GIMP to edit it, then did I take a photo or just direct my camera at something and then steer my computer to make changes to it? After all, it's not me drawing each pixel manually. It's vibe-photoed!

Tragic day in AI.

What exactly is broken with copy-paste? I've been using Zed on Wayland (official install, not flatpak and not terra) and had no issues with this.

I guess it does make their communication feel a bit more personal, especially since they usually don't use corporate language on personal twitter accounts.

Truly a terrible day in AI. We shall all rise against the 5-hour limit.

Also some Sol token usage optimizations are coming, as described in the tweet. Seems like OpenAI is really trying to remain the leader in token efficiency.

Lower barrier to entry = more crap flooding the market

I'd say it's more like going to a Ferrari dealership and they tell you they will build a car for you and bill you per gram of parts used. It might also not work. And they might also not build it ever, really — but they will bill you for any attempts to build it.

I am not sure if pricing actually reflects inference costs

Yes that's true, although calculating the exact discount you get using a subscription could be difficult, especially since the labs often don't state their actual limits and just tell you you get "more usage" or "even more usage" or "5x the usage" (but how much exactly is x?).

It is clear we are getting a very meaningful discount by using a subscription though. I once checked how much would I pay for Codex after about 3 months of non-daily use if I only bought "credits". Couldn't believe it when it came out to be north of $1000 (I paid them $20/month in that period).

Well yeah this benchmark I've quoted is quite large, I believe they use thousands of diverse tasks, so this average cost per task could be a more accurate representation of how expensive each model actually is to run.

I also don't see much of a quality difference switching between thinking levels while using those models as agents.

It's very impractical.

- you get a slower first load (cannot progressively fetch resources as they're needed) - can't reuse a stylesheet, script or image on a different page (each has to have their own copy) - can't cache commonly used files - can't make granular changes to specific parts of the code. user has to reload everything each time. - can't set a proper content security policy

And many more! It's cool for a tiny demo but for anything serious you wouldn't want a single (extremely ugly) HTML file.

From the linked Hollywood reporter article:

"...the AI company exits the video generation business."

"OpenAI, led by CEO Sam Altman, is not getting out of the AI video business [...], of course... "

I hate journalism.

There are no tricks but you can manage it to an extent.

- Find activities you can do *outside*, regularly, preferably with a group. Things that worked for me: Learn japanese (group lessons only! Invite everyone for a beer afterwards), find a literature club, go swimming every other day, go cycling when bored.

- Reach out to friends and acquaintances you didn't speak to in a while. Some won't reply, others will be delighted you reached out. Most people meet their s/o through mutual acquaintances.

- Go on websites like Meetup (or whatever event board is most popular on your area) and go to every event you can manage. If you're not the type to strike conversations with random strangers then find events looking for speakers or hosts and people will come up to you themselves.

It's *very normal* to feel empty when a large part of your life is suddenly not there. I couldn't get rid of that feeling at any point when I was alone. Fill in the empty time and space, try finding someone else in a while but don't rush it—otherwise you risk yet another breakup.