[0] https://amplifying.ai/research/claude-code-picks/report

Something happened, we won't say what, but it was severe enough to notify law enforcement. What floors me is the only actionable advice is to "review environment variables". What should a customer even do with that advice? Make sure the variable are still there? How would you know if any of them were exposed or leaked?

The advice should be to IMMEDIATELY rotate all passwords, access tokens, and any sensitive information shared with Vercel. And then begin to audit access logs, customer data, etc, for unusual activity.

The only reason to dramatically overpay for the hosting resources they provide is because you expect them to expertly manage security and stability.

I know there is a huge fog of uncertainly in the early stages of an incident, but it spooks me how intentionally vague they seem to be here about what happened and who has been impacted.

The writer went to SF for a few days and went to one party where a group of friends were into peptides. From the article, they were also particularly terrible people. Just read this quote:

> “They change your personality, it’s literally made me less shallow knowing that we can just looksmax you.” “Ugliness is just a choice now.” “I shot up a twink with ozempic who did not need to lose any weight.”

I can’t believe I have to say this, but if someone is bragging to you about injecting weight loss drugs into another person who shouldn’t be taking weight loss drugs, your response shouldn’t be “lol how quirky”. You should recognize that they are a bad person. In my experience the drug enthusiasts who brag about getting other people started on their drugs are bad news, but the ones who brag about introducing to their drugs to people who clearly should not be taking those drugs are the worst variety.

These people always exist. Go back a few years and they might be talking about nootropics or “research chemical” drugs that are analogs of methamphetamine or MDMA. Go back further and they might be bragging about doing steroids and importing testosterone from gray market sources. Go back before that and they’d be bragging about all the Modafinil they’re taking.

The thing about drug user bubbles like this is that when you’re talking to them you’d be convinced that everyone is doing what they’re doing: Taking the latest on-trend drugs in large amounts and one-upping each other on dose, stories, or drug-fueled adventures.

What’s not talked about is the long-term consequences of falling into these groups where excessive drug self-experimentation is normalized. The party doesn’t last forever and the mindset of being able to endlessly adjust your body and/or your mood with drugs starts to turn dark after the early years where hubris makes users feel like they’ve found the secret to better living through chemistry.

If you’ve encountered groups like this you’ve also seen how the “everyone is doing it” mentality becomes embedded in their minds. That doesn’t mean everyone is importing various Chinese peptides and injecting them for “looksmaxxing” and whatever these people were on about about the “peptide party”. These are just garden variety young drug users riding the latest trend

EDIT: I replaced one instance of the word ‘journalism’ with ‘writing’ because it was becoming a pedantic distraction in the comments.

Since the game uses a 256 color palette, it was only necessary to update a few bytes of data (3x256) instead of redrawing the whole screen, so the effect was quick.

I also used this trick when the game stalled due to missing network packets from other players. Initially the game would still be responsive when no messages were received so that you could still interact and send commands. After a few seconds the game would go into paused state with grayscale screen to signify the player that things were stuck. Then several seconds after that a dialog box would show allowing a player to quit the game.

This was much less disruptive than displaying a dialog box immediately on network stall.

The UAE doesn't have a self-advancement culture, it's a capital-backed monarchy that imports pretty much all of its research and production; in other words it piggy-backs on the knowledge produced in other societies. There is no advancement through dialog in the country itself.

Is the charge, which I think kind of speaks for itself. Full on: "You embarrassed us, straight to jail."

In most of the world such photos would be deemed of public interest and shared by the media then we'd reflect on if our routing is safe/correct and make proportional changes for safety. Not a big deal, nobody is fired, life moves on.

I feel like actions like this are going to hurt the UAE themselves, because how can you improve if there is no dialog? No information to even start a dialog? A lot of hard conversations are NOT going to be had because I guess it is a state secret?

There was also almost nothing to read, so when my monthly issue of Byte appeared (2-3 months later than most people would receive it), I devoured that thing. I would read it literally cover to cover, including all those ads, several times.

I wasn't (then) working in IT, so a lot of the content (like Steve Ciarcia's Circuit Cellar) went way over my head but it didn't matter, I read it anyway, often by the light of my kerosene lantern. I learned a huge amount: object-oriented programming, this new thing called the Internet (capitalized back then, and before the WWW), and how Jerry Pournelle was a self-important jerk (but boy, did I envy the toys he got to play with!).

This was the age of big, fold-out Gateway 2000 ads, 20MB hard drives, and Turbo Pascal kicking other compilers' butts.

I would read the magazine, then write out programs (in BASIC, the only language I had learned at that point). On my monthly trips to the capital city I would go to a local NGO and in exchange for helping with their IT issues they would let me play (i.e type out my programs and try to get them working) on their computers.

It always surprised me how few games had that feature - though a few important ones, like StarCraft, did - and it only became rarer over the years.

Dad here. Maybe…it’s the lack of sleep? Involved fathers tend to have less sleep.

We already had AI proof education.

Then I looked at the usage and it said I had used 95% of my Claude design usage for the week!

This isn't a real tool. This is a plaything, if that's what they're providing as examples.

[0] https://storybook.js.org

Here is a comparison for 4.5, 4.6 and 4.7 (Output Tokens section):

https://artificialanalysis.ai/?models=claude-opus-4-7%2Cclau...

4.7 comes out slightly cheaper than 4.6. But 4.5 is about half the cost:

https://artificialanalysis.ai/?models=claude-opus-4-7%2Cclau...

Notably the cost of reasoning has been cut almost in half from 4.6 to 4.7.

I'm not sure what that looks like for most people's workloads, i.e. what the cost breakdown looks like for Claude Code. I expect it's heavy on both input and reasoning, so I don't know how that balances out, now that input is more expensive and reasoning is cheaper.

On reasoning-heavy tasks, it might be cheaper. On tasks which don't require much reasoning, it's probably more expensive. (But for those, I would use Codex anyway ;)

No wonder they hallucinate :)

One of two main reasons why I'm wary of LLMs. The other is fear of skill atrophy. These two problems compound. Skill atrophy is less bad if the replacement for the previous skill does not depend on a potentially less-than-friendly party.

We'll be keeping an eye on open models (of which we already make good use of). I think that's the way forward. Actually it would be great if everybody would put more focus on open models, perhaps we can come up with something like the "linux/postgres/git/http/etc" of the LLMs: something we all can benefit from while it not being monopolized by a single billionarie company. Wouldn't it be nice if we don't need to pay for tokens? Paying for infra (servers, electricity) is already expensive enough

This is got to be a huge factor. Making everyone pay for "free parking" through inefficient use of space is such a waste. I strongly recommend everyone to read Donald Shoup's "The High Price of Free Parking".

This is the most important paragraph in the article. It can’t be overstated how ingenious Japan’s system of zoning is and how much this has benefitted their society in ways we can only dream about here in the West.

Rabin Fingerprinting is one of my favorites of his contributions. It's a "rolling hash" that allows you to quickly compute a 32-bit (or larger) hash at *every* byte offset of a file. It is used most notably to do file block matching/deduplication when those matching blocks can be at any offset. It's tragically underappreciated.

I've been meaning to write up a tutorial as part of my Galois Field series. Someday..

Thank you again!

The thing with computational geometry is, that its usually someone else's geometry, i.e you have no control over its quality or intention. In other words, whether two points or planes or lines actually align or align within 1e-4 is no longer really mathematically interesting because its all about the intention of the user: does the user think these planes overlap?.

This is why most geometry kernels (see open cascade) sport things like "fuzzy boolean operations" [0]) that lean into epsilons. These epsilons mask the error-prone supply chain of these meshes that arrive in your program by allowing some tolerance.

Finally, the remark "There are many ways of solving this problem" is also overly reductive, everyone reading here should really understand that this is a topic that is being actively researched right now in 2026, hence there are currently no blessed solutions to this problem, otherwise this research would not be needed. Even more so, to some extent this problem is fundamentally unsolvable depending on what you mean by "solvable", because your input is inexact not all geometrical operations are topologically valid, hence an "exact" or let alone "correct along some dimension" result cannot be achieved for all (combination of) inputs.

[0] https://dev.opencascade.org/content/fuzzy-boolean-operations

It's defined as the difference between 1.0 and the smallest number larger than 1.0. More usefully, it's the spacing between adjacent representable float numbers in the range 1.0 to 2.0.

Because floats get less precise at every integer power of two, it's impossible for two numbers greater than or equal to 2.0 to be epsilon apart. The spacing between 2.0 and the next larger number is 2*epsilon.

That means `abs(a - b) <= epsilon` is equivalent to `a == b` for any a or b greater than or equal to 2.0. And if you use `<` then the limit will be 1.0 instead.

Epsilon is the wrong tool for the job in 99.9% of cases.