idreyn

I am not sure which rail right-of-ways in the Bay Area could plausibly be justified for AV lane deconversion. Maybe the ailing VTA light rail?

But an "AV lane" will very quickly converge on the traffic characteristics of a regular road.

Here are the two I was thinking of:

- https://news.ycombinator.com/item?id=49767937

- https://news.ycombinator.com/item?id=49784816

These last few days, there has been a small stream of blogposts on HN expressing similar things, and I have enjoyed them all.

My question recently has been how to broach this subject with colleagues who really enjoy producing prose with AI. There is not yet a better cultural shorthand for this sort of thing than "slop" which is a harsh-sounding word and itself sort of a thought-terminating cliché. "I don't want to read what you didn't write" is maybe closer — but it needs a pithier and somewhat more encouraging encapsulation, like "I want to hear it from you".

Has anyone had good experiences setting up professional boundaries or team norms around AI-written docs?

I'm rooting for this general class of technology as a platform for sensory augmentation for the blind. I don't know exactly what sonic encoding of spatial information is exactly ideal — I suspect it's something echo-like, only slowed-down — but it's going to be a hell of a lot easier to develop it and support it on commodity hardware. Please no Meta login, though.

Grok is that you?

> He argued that Modern Chinese has become “lazy” by forgetting how to use its own verbs. instead of “researching” (研究, yanjiu), speakers “conduct research” (进行研究, jinxing yanjiu)

I can't help but think of this classic essay about Java OOP: https://steve-yegge.blogspot.com/2006/03/execution-in-kingdo...

I think the more mainstream understanding of transhumanism is closer to "post-genics" than eugenics.

Flash created a medium. The particular genius of the authoring tool gave rise to a whole style of animation and game and thing-in-between that only existed in its time and could have only been created with the tool at hand. Software should aspire to this.

did we read the same post? none of what was mentioned was even in the realm of entrepreneurship or anything ethically hairy.

My word count has hovered around 100k for most of my three years of writing and revising. This does sometimes run up against limits on Claude (or recently, with Opus 4.5, compaction) but in the past the whole thing has fit just fine as a plain text file.

Yes. I am a novelist and I noticed a step change in what was possible here around Claude Sonnet 3.7 in terms of being able to analyze my own unpublished work for theme, implicit motivations, subtext, etc -- without having any pre-digested analysis of the work in its training data.

NYC here, our three disconnected commuter rail systems would all like a word :)

These maps are such a cool resource, thanks for sharing!

"The future is already here – it's just not evenly distributed." - William Gibson

It is to render untrusted (user-generated) HTML without letting them slip in markup like script tags that could harm other users.

Yes, it is. And yes, except that it wasn't. A SWF is about building common wealth inside the systems that finance capital built (in the same way that the 401k replaced the pension) rather than turning back the clock on them. How you acquire those assets can vary wildly:

- Maybe you just decide to invest some public money

- Maybe you have some natural resources that are collective-by-default (minerals wealth on public land)

- Maybe there's a bailout of an industry that is financially broken but has become too big to fail cough and the government presses its leverage

- Maybe a president just wakes up and decides that he wants the government to own 10% of Intel, and makes that deal happen on favorable terms.

This sounds a lot like a sovereign wealth fund. The government obtains fractional ownership over large enterprises (this can happen through market mechanisms or populist strongarming — choose your own adventure) and pours the profits on these investments into the social safety net or even citizens' dividends.

For this to work at scale domestically, the fund would need to be a double-digit percentage of the market cap of the entire US economy. It would be a pretty drastic departure from the way we do things now. There would be downsides: market distortions and fraud and capital flight.

But in my mind it would be a solution to the problem of wealth pooling up in the AI economy, and probably also a balm for the "pyramid scheme" aspect of Social Security which captures economic growth through payroll taxes (more people making more money, year on year) in a century where we expect the national population to peak and decline.

Pick your poison, I guess, but I want to see more discussion of this idea in the Overton window.

I would like a non-native speaker to weigh in, but I gloss it as "there are ONLY a few and this is ONE OF them" and have never found that confusing or contradictory.

It's a stereotype of his political opponents as adrift in selfishness, hedonism, and "gender stuff".

When you find your thought patterns flowing through tropes this way it is a good sign to consider logging off for a long while, as the author is doing. I wish him peace and perspective.

This is clearly done by AI and I am disinclined to draw any conclusions from the anecdotes or data here because it all feels "synthesized".

On one hand, I wonder if a gradual transition would work. Spend enough time over the years mirroring your conscious patterns onto a computational substrate, and they might get used to the lay of the land, the loss of old senses and the appearance of new ones. There might not be an ultimate "stepping in", but something like you might be able to outlive you, on a substrate that it feels happy and comfortable on.

On the other hand, the idea of "simulating your consciousness" raises questions beyond just cognition or personality. A mechanistically perfect simulation of your brain might not be conscious at all. Spooky stuff.

Been thinking about this. Do you have a snippet you can share?

this is very smooth and scratches an itch I didn't know I had -- to use a windowed OS on my phone

was also looking for some KEST GAK stickers

I have no affiliation with TOC; just thought this was neat.

Lovely. I've wondered about this for years but never knew quite how to frame a query to learn more.

WEI acts as proof that "this is a browser", not "this is a human". But browsers can be automated with tools like Selenium. I'd guess that with the advent of complicated, JS-based captchas, browsers under automation are already the major battleground between serious scrapers and anti-bot tools.

I also don't understand how WEI does much to prevent a motivated user from faking requests. If you have Chrome running on your machine it's not gonna be too hard to extract a signed WEI token from its execution, one way or another, and pass that along with your Python script.

It looks like it basically gives Google another tool to constrain users' choices.

I wrote to the FTC: https://gist.github.com/idreyn/a37e4249b326243e1714b14780717...

Hooks let you colocate logic that, in class components, would need to be split across multiple lifecycle methods. In practice you need to remove the event listener when the component unmounts. You can get reasonably close to a hooksy API for doing this here:

    class MyComponent {
        componentWillMount() {
            this.stopWatchingViewport = watchViewPort((x, y) => {
                this.setState({ something: x + y });
            });
        }

        componentWillUnmount() {
            this.stopWatchingViewport();
        }
    }

    watchViewPort(callback) {
        const onResize = (event) => {
            // get x and y
            callback(x, y);
        };
        addEventListener("resize", onResize);
        return () => removeEventListener("resize", onResize);
    }

But being able to slice up logic by functionality, rather than by lifecycle event, gets gradually nicer as you have more of it.

I have a hook that I copy around between projects called `useViewport()`. All it does is watch the current viewport size and updates its value when that changes. Because it's a hook, I can call this function from any component body:

  const { width, height}  = useViewport();
  return <div>Viewport: {width} {height}</div>

Zooming in, this is just built from a useEffect() to bind event listeners, and and a useState() to hold the current value. Zooming out, it would be easy to write a useBreakpoint() hook that mostly just calls useViewport(). Hooks compose.

Before hooks, there weren't great patterns for isolating units of stateful logic from the presentational side of React. A really common pattern was to use "render props" where a component keeps some internal state and passes it into a function-as-child:

  <ViewportWatcher>
    {({ width, height}) => <div>Viewport: {width} {height}</div>}
  </ViewportWatcher>

You'd see these provider-components nested three or four layers deep, and it got ugly. Hooks solved this.

I think that's basically right. They've been weaving "high-status outdoor adventuring" into their brand for a long time — think about the codenames High Sierra and Mavericks, or going to an Apple store and seeing a ski trip under planning in the sample iMessages.

A big part of the way Apple signals their products as clean, healthy, and high-status is by depicting themselves as exactly that sort of people.

(Not to imply that this is particularly disingenuous, but they know exactly what they're doing)