swiftcoder

> It's where influential folks post. Whether you like it or not.

It's where a bunch of rich (and often very right-leaning) assholes post. I'm not clear that correlates with "influential" outside of very specific circles - certainly all the thinkers and doers I used to follow on twitter left long since.

> Even Zuckerberg tweets.

Having worked for the man some time ago, I'm pretty sure a bowl of oatmeal has more insightful things to say

I mean, the rest of us just stopped going there. The interesting tech discussions long since left twitter, and if you are jonesing for a fix there is always bluesky or mastodon

> if you spent a lot of time on X you will likely remember a particular one: everyone seemed to be suddenly becoming fans of Adolf Hitler... I did not leave X the platform, but I did leave X the interface

You know, it really is entirely optional to hang out at the Nazi bar.

> A phone connected to cell is an always-on tracking device

A phone connected to wifi is also an always-on tracking device. You can't really escape this without removing the "phone" part of the equation

I don't see how the cost of inferencing existing models can become prohibitive. Multiple providers are selling DeepSeek 4.1 Flash tokens at just above cost, and seem to be doing so profitably.

What we do potentially lose out on are the next few generations of models in the case of a market correction, as the big labs won't be able to afford training anymore.

> AI results from search engines don't generate the click-through revenue that search ads did

The question in my mind, is what were all the Google execs smoking when they decided to kill the golden goose?

It's not like this is a non-obvious outcome of sending all your most valuable searches to an in-house AI answer instead of an ad click...

> On the other hand kings had free-range organic grown meat prepared by personal cook anytime they wanted which is a true luxury even

There is no refrigeration, so "fresh" meat has to be slaughtered the same day. Kings tended to live in populated areas, so wild game would be scarce. Majority of meat would have been pork kept in pens (not free-range in the modern sense), or mutton/goat (free range, but typically older and tougher). At the same time, even a king likely wasn't slaughtering 365 pigs/year just to have fresh meat every day - lots of dried, smoked, or even salted meat in the diet...

That's a really concrete demonstration of gerrymandering. Assign it in schools!

I'm actually unsure whether this is satire or not, which says volumes about the current moment

> human negligence and the system doing exactly what it’s supposed to do

Negligence is not limited to the single human who issued the order. it's pretty unlikely that their order could have caused this without a significant amount of other negligence that eroded the safeguards. Nevertheless, the negligence is human, not machine in nature.

> that you deterministically and provable know where a train is

You are the one who framed this conversation as it not being solved how to deterministically know where a train is. Everyone else in here has been kindly suggest that if perfect information is unobtainable, then we should work with what we have

> Yeah, but how do you distinguish that?

By enumerating the tracks?

I feel like you are intentionally trying to turn this into a spherical cow situation by insisting on solutions that work in an information vacuum. All of this sort of problem in the real world is solved by sensor fusion from multiple data sources.

Trains are captive to a particular track - they can only change tracks at specific pre-determined junctions. With a pretty minimal sensor feed from each junction we can statically determine which track segment a given train is on. With GPS we can shrink that segment length down to a few kilometres. By sharing data persistently every single train can validate that central command-and-control is not about to run it into another train from behind...

There cases which are genuinely challenging engineering problems (i.e. derailments that impinge on adjacent tracks), but this trivial case of bumper cars is not one of them.

> But why would you transmit all this information to every train, instead of just computing the collisions in a central node?

Because central nodes can fail (as they did here, both human and machine), and providing this redundant data allows both the computer and the human-in-the-loop on the fast-moving chunk of steel to independently validate the decisions coming from the central node.

> You regularly pass other trains within <2m at high speeds, so linear offset is not all you need.

In this collision, both trains were on the same track. So while we obviously care about adjacency issues in the general case, this specific failure mode can be detected purely via a linear offset.

These high speed trains are pretty recent. Are we really not upgrading safety systems past the 1930's when upgrading a line from ~80kmph to ~300kmph?

> Trains are still not able to assess properly where they are, even using satellites and inside train sensors. Of course there is research on that, but it's still an unsolved problem

Throwing up our hands about how hard a problem this is seems like a ridiculous position to stake. If we can design self-driving cars, that operate safely in far less controlled environments, we can figure out the linear offset of a large vehicle traveling along a physical rail running along a fixed pre-determined path.