SamBam

If they make it hard enough to interact with the actual website, we'll have to use AI!

Nice!

Needs a chat space to say "gg" at the end. It could even be just a button that does just that, if you're worried about unmoderated chat. Competitive games feel unfinished without it!

I was in second place for the first three rounds and then swept to the front on the last round.

Also note that Goodier's post that kicked off Gruber's post now says it had a section about "an astronomy app I am working on that began as an astrology app," [1] but the older version of the article clearly implied that the author was baffled as to why Apple would ever think it related to Tarot cards. [2]

1. https://blog.terrygodier.com/2026/08/07/browsers-have-standa...

2. https://web.archive.org/web/20260807141155/https://blog.terr...

Not really, because the capabilities this announcement advertises is exactly three capability some people want to defend against (and others want).

It might be more on par with a for-profit fire department showing how -- oops! -- easily buildings catch on fire these days.

Should the leader board be cleared of the folks with 10,000 landings in 1 second?

It looks like the true leader right now is 90 landings in 7:29.

Very fun. Got a few minutes in. Is there a strategy for avoiding hitting the planes driving slowly across the runway for takeoff? It seems very hard to also pay attention to them when they have so little lead time.

Yes, that one worked fine for me as well.

Many people don't realize that regular spaghetti actually is pretty high in protein. All that gluten.

Why would AI say that, when nearly every piece of training data is ever been fed would use milliseconds? I think much more likely that this is how the author thinks of it.

3 hundreds of a second == 33 Hz is very clean in my human brain.

I think the objection about AI-generated cards, or decks found online, is that the context of the card creation is lost.

What made you write the card? How did you find the answer/translation/whatever? What was all the learning around the card that the card is supposed to represent?

If you're just doing simple word translations into a new language maybe that's ok. But if you're trying to learn a concept in organic chemistry or SQL, then you're more likely to memorize the card itself than learn the concept.

The AI in the article isn't an LLM.

> Hacker culture sees programming as a conversation with a machine as it runs.

I wrote about this a fair bit as a party of my master's thesis, about tinkering as being a "conversation" with materials. Hacking can absolutely be looked at in this framework.

The philosopher Donald Schön I think was the first person to formulate tinkering in that way. The process of engaging with materials -- whether it's a broken motor you're curious about or a tune you're plucking out as a complete novice or some code on Scratch -- involves asking questions of the materials, learning to hear answers, and noticing when the materials themselves pose questions. It's a really nice way of looking at things.

I can't imagine how much amazing and important literature you'd miss if you were snobby enough to think that you could only read things in their original language.

I'm so glad I get to read the Russians and Kafka and Calvino and Murakami and Camus and Marquez and Homer and Plato and, heck, the Bible.

I do know the feeling. I struggled through the start of My Brilliant Friend because I ought to read it in Italian, because I speak it pretty well. So then I didn't read it for years. Finally I just read it in English and enjoyed myself.

I was confused when I first read it as well, but the implication is that the higher reward-seeking risk tolerance can actually be more rational than the well-studied loss-aversion bias that the authors mention.

Looking into this more, studies have found that we tend to rate the possible loss of $100 twice as painfully as the pleasure from the possible gain of $100. This can lead to irrational behaviors.

Increasing the weight we give to potential rewards is not necessarily a bad thing.

I think this can help explain the "calming of the nerves" that slow breathing promotes. If you need to speak in public and your heart is racing and you're shaking, this is an irrational reaction to what ought to be a very safe situation. By focusing more on the rewards (the acclaim for a good speech or whatever) and less on the imagined risks, you can calm down and speak naturally.

> The important catch

> The favicon doesn't actually contain the whole website itself.

This is the kind of thing that is extremely idiomatic LLM speak. There's nothing particularly wrong about it per se, but it just makes everyone who is familiar with LLMs say "oh, it's written by an AI" and it just becomes disappointing.

A large percent of espresso in the US is drunk iced. Iced Americanos, iced lattes, iced shaken espressos.

If you don't need to heat the water before cooling it down it would be an enegy savings.

> For the espresso samples, participants could not reliably tell the traditional and ultrasonic versions apart. There were no significant differences in aroma, flavour, bitterness or overall liking.

Nah. I have a Roku 3 stick and a brand-new $700 projector with Google TV. The Roku 3 is light-years ahead in terms of speed and UI ergonomics over the Google machine. And both are better than the smart TVs I've used.

But I fear this need means this time is ending, and we'll only be left with crap.

I'm a teacher. Our kids are using Chromebooks, and we want them to do research projects, but we don't want them using AI. Unfortunately with Google, it's impossible not to get AI answers. As far as we know, there's no clean way to filter out the Gemini garbage. I wonder if Kagi has considered making an educational pricing model, and letting our kids get away from AI.

Of course, the "Chromebook" part of my statement above is probably a hint that getting rid of Gemini won't be as simple as changing the default search engine...

I feel like unresponsive controls must have been a platform or hardware issue. On my (...Dell with Windows 3.1?) the controls were perfectly responsive, and they absolutely had to be. So many things relied on being able to, say, run and switch directions on a dime so that only your toe touched the platform that would then fall and break.

The tutorial made it seem a little too much like there is only one speed that would keep us in orbit. Any slower and we'd crash, any faster and we'd leave.

In fact, though, if you've ever played any game with orbiting mechanics you'd see that it's extremely difficult to get out of orbit if you're in orbit. Going faster simply increases the size of your orbit, and going slower simply shrinks it.

Note that no space program has ever managed (or tried) to send an object into the sun. We're already starting off with such a high orbital velocity, 30km/s, that we'd need to send a rocket backwards at nearly that speed just to slow it down enough to make it crash into the sun. That would require massively more energy than anything we've ever done before.

Everything you're saying is right, but I'm not seeing what's wrong with step 14. Did they edit it?

> Earth turns once every 23 h 56 min (one sidereal day) about an axis tilted 23.4° (the blue line). That spin gives us day and night; the tilt gives us the seasons.

Nothing in step 14 to me implies s procession of the axis.

Huh, perfectly snappy on my Firefox on Android.

> wtaf? Do you really think this is the reality?

Of course. American varsity soccer players play ~5 times a week for four months a year, August through November.

February through April is the Spring season and varsity teams are forbidden from playing too much. D1 and D2 teams can only train 8 hours per week, of which at most 4 hours per week may be coached!

And then 4 months of the year are off-season and there is no structured training at all!

This is a ridiculously small amount of time compared to a 15 year old Spanish kid playing in a farm team.

And they're doing this part-time stuff until they're 22 or 23. A typical European professional player would have been playing full time since they were 16.

> college soccer is essentially another international semi-pro league

... except no where near at the level of European farm teams.

Basketball is the only sport in the US that has a similar level of hundreds of hours a year of unstructured play, so it makes sense that the US is great at it.

Hockey, who is the US competing against? Canada. No other country in the world takes it particularly seriously, so of course the US is one of the top two countries... (I often tell each of my two kids that they're one of my top two kids.)

So leaving aside whether 5-6 times is "a few," the bigger issue is the length of the season.

Varsity soccer season in the US is usually just four months, August-November.

Spring season (with no games) is February-April. During that season, NCAA places strict limitations on how often teams can practice: Division I and II teams are allowed only up to 8 hours per week, with just 4 of those being coach-supervised! [1]

Finally there is no organized playing for all of January, May, June and July.

So even for a player in a D1 team, they are training much, much less of the year than a 15-year-old on a farm team in Europe.

1. https://ballatyourfeet.com/when-is-college-soccer-season-fal...

I don't think the article really tried to answer the question, though maybe that wasn't its intent and the author was genuinely asking.

I think an answer would need to look at the difference in how kids and teens play soccer in the US vs other countries.

In the US soccer is mostly a younger kids' sport, and is generally highly structured, with kids playing on teams once or twice a week. Compare to Europe, where many boys are playing once or twice every day, in an unstructured format, during recess and after school.

Starting from a young age, Europeans who show talent are getting drafted into soccer academies before they're 10, greatly increasing the amount of competitive play. But this is on top of the everyday soccer they're playing.

For a US kid, soccer is typically "pay to play." A local league costs money. A private high school with a good program costs money. In Europe, beyond (again) the continuous unstructured play, the academies and farm teams are free.

Finally, a good European player doesn't usually head to college. They may be playing for a serious club team at 17 or 18.

Meanwhile, a gifted US soccer player heads to college (maybe on a scholarship but maybe not--again, pay to play), plays for the varsity team a few times a week during the season, and four years later might get on one of the relatively few club teams.

Same order of magnitude.

User should be required to explain the situation to an older and a younger family member, and get permission from both of them.

I wonder how accurate a virtual model could be made of this, which could be iterated on millions of times faster.

My first project as a research assistant in AI was doing evolutionary algorithms on Khepera robots, which had a virtual Java implementation. We were able to evolve some pretty cool behaviors, although I don't know what would have happened if we had uploaded them into a physical Khepera robot.