miki123211

I have said this many times, and will probably have to say it many more: Stablecoins are nothing but a face-saving measure for the US government.

It is politically impossible to go back on KYC/AML measures in the traditional financial system, even if some of them were perhaps a bit too ill-advised. It's much easier to say "this is shiny new tech, very different from that old tech over there; we support innovation so we won't impose KYC requirements on it", even if there's no logical reason to do that.

As it stands, stablecoins let you make a keypair in the privacy of your own home (or your own agent sandbox, as it were) and immediately start transacting. They're allowed-by-default instead of blocked-by-default, with the US government still able to block you if they have a specific reason. No traditional bank I know of, in any country, works like this. This is a valuable capability to have, technologically, even if there's no technical reason why a Postgres database and a REST API couldn't provide the same service, but faster and more easily.

Because these capabilities are dual-use.

You want canvas to work for web games and design / drawing sites. Yet. minuscule differences in how canvas is implemented let other sites fingerprint you. The only way to fix this would be to have each browser be based on the exact same codebase and web engine, and do you really want that?

remember about aria-hidden or equivalent, otherwise you are making the text visible to screen reader users.

I have read many snide comments and personal ID numbers that way, mostly in public and public-ish documents.

What makes this even more confusing is reasoning effort.

You have Luna, Terra, Sol and Astra, each with 4+ reasoning levels. That's ~20 options to choose from.

On a repetitive task like customer support or something, where you have decent benchmarks, you can plot cost versus performance, but you can't really do that on coding.

Machine learning is more like biology, physics or law than like CS or math.

In the software engineering world, we're accustomed to the reality that to do great work, all you need is a sharp mind and a cheap laptop. Better tools help, but aren't really a necessity. You could probably become a top-notch Linux kernel engineer with nothing but dedication and self-study, a $300 laptop and a shabby apartment with internet access.

You just can't really learn ML that way, just like you can't do it with banking, professional trading or medicine.

What I wish people understood about privacy is that the less tech adjacent a company is, the worse their privacy record.

Meta and Google will never sell your data. They will collect it and use it for targeted advertising for sure, but they consider it far too valuable and useful to sell. This is because they have the tech and expertise to use their data well.

A grocery store or a gas station chain, though? They're clueless about tech, don't know what to do with the data, and have to aggressively compete on prices. If some savvy data broker comes to them, offering them a way to essentially gain market share by pushing prices lower than the competition and selling your data to make up the loss? That's an offer they can't refuse.

Hardware manufacturers are largely on the grocery store side; they're not data-savvy and they want to make things cheaper than the competition, so they're almost forced to do these shenanigans to win the competitive rat race.

If you want to meaningfully increase privacy, don't quit Meta, quit Target, Visa and LG.

We used to solve this problem by embedding experts in government bodies.

We can't do that any more; the people who support tech regulation are so cartoonishly anti-corporate that any suggestion of embedding somebody who understands tech will be met with accusations of lobbying and a careful review of that person's employment history.

It's obvious to pretty much everybody that regulating banks, railroads or phone companies requires people who have previously worked for banks, railroads and phone companies, and have a good understanding of how such organizations work on the inside. We, generally, don't do that with tech.

The major other reason is that it is free, and most people are money-constrained.

Personalized advertising is a progressive tax. Because the rich spend more and the poor spend less, Meta et al makes more money from the rich than it makes from the poor, while offering the same service to everybody else. As a result, we get the educated rich upper-middle class complaining about advertising, while everybody else is happy about not having to pay their long-distance charges any more.

Hardware companies (and Japanese hardware companies in particular) don't care for or understand the rules that govern software.

Neither do governments, incidentally. They're okay-ish at skyscrapers and bridges, but software is just beyond them.

Agent parallelism is cheap in the same way tapwater is cheap — that is, you pay exactly for what you use at a given moment, no more and no less.

Humans are a lot more expensive, it takes ~a year to find, interview, hire, onboard and offboard an employee, while an agent can be spun up for 10 seconds, do its work and vanish back into the ether.

Agents are lambda functions where humans are bare-metal servers that you have to rack yourself.

And it's also AGPL, which is a legal gray area and a headache in many places, even those that technically aren't violating the AGPL.

I am a Python programmer. I love Astral; I think Ruff is great and uv is life-changing, but I don't actually like ty that much.

I want my type checker to check types and to scream at me when it can't. I don't want it to do "gradual typing", if it is unsure about what type something is, I want that to be a failure instead of it modeling the unsureness and not showing potential errors. We live in the age of agents, having to add an extra annotation here and there is not so much of a problem any more.

Meta's Pyrefly largely does what I want. I don't quite love Pyrefly either; I don't like the fact that it edge-cases and gives monopoly status to some popular packages (Pydantic, Django, attrs) instead of offering generalized mechanisms that less popular packages could also utilize, but purely on a type checking front, it's excellent.

TBH, your average AI agent in 2026 is far, far more capable than anybody in tier1 or tier2.

The reason you need heavy scripting in those tiers is because the people you can afford to hire there are incapable of literally anything else. Keep in mind that your average HN user isn't your average customer support caller. The average customer support caller just wants to talk to somebody (yes I'm serious, this happens far more than you think), or has a routine issue which is trivial to fix but which they're nevertheless unable to fix themselves, think "router got unplugged, there's no internet, but the customer can't figure out why".

A well-designed AI system can be far, far more effective and pleasant than traditional CS, and a well-designed AI system scales well, so it's actually worth spending money on. AI turns good CS from a cost center to a profit center.

I wonder if it's as good at "soft skills" as previous Gemini models were. Google was typically better in that department in my experience than either OpenAI or Anthropic, whose frontier models are focused on coding and verifiable task at the expense of all else.

What really impresses me about these AI visualisations is that they're often (but by no means always) fully screen reader accessible, with proper keyboard and ARIA support. Those by Ciechanowski very much aren't, and even vibe-coded slop accessibility is better than no accessibility at all.

Accessibility with AI is hit-and-miss; if you don't tell it to make things accessible, it either will or it won't, and then it'll just continue with whatever trajectory it chose. If you tell it to do accessibility initially, it'll see its own ARIA later and continue in a similar vein.

From very anecdotal experience, I find OpenAI models more likely to consider accessibility for greenfield projects, even when not asked to, but this might just be an accident of statistics.