Elon just made a pertinent statement:
https://futurism.com/future-so...

That's not why. Waymo is testing in Manhattan and Downtown Brooklyn. NYC by the way has also recently legalized jaywalking.

The idea that machines don't / won't beat humans at continuous vigilance and precise movement doesn't make much sense to me, since machines are great at that. The safety issue already favors automation and the gap will only grow. (More specifically, safety already favors certain self-driving implementations, like Waymo... obviously in general, "automation" can also be total crap if done poorly).

I know we are only relatively early in the development and adoption of the technology, but I sure can't see any reason to doubt the outcome.

Not only are nuclear reactors expensive, the build time is twice what the major nuclear construction companies would say it will take.

Hinckley Point C in the UK is still being constructed - it's a 3.2GW plant, it is costing twice the cost as planned (in 2015 prices) and is going to be 4 years late. In the time it's taken we've installed 28GW of wind and solar capacity.

That's going far past anything I would know; I'm software, not hardware and it was a long time ago. All I remember is that there were a pair of uprights that helped hold the chip up and it went from the motherboard almost to the top of the case.

This reminds me of something that was done back in the (I think) 90s for one of the Pentium chips. Instead of it lying flat on the motherboard it had all of its connectors along one edge and stood upright on that edge in a special mount that kept it upright so that all of it was exposed to the air and didn't need a heat sink or special fan. Yes, it had its drawbacks, mostly that it couldn't be used in a laptop and needed a tall case, but it worked and worked well. I know, because I used one for several years back then and only replaced it to upgrade.

Made me curious. Here's a graph of their headcount over time.

Some possibilities:
-The agent buys the wrong thing and Amazon sees a substantially higher rate of returns or other bad customer feedback
-The agent buys one thing despite Amazon search results trying to push a different option
-Amazon's upsell for "you may also like" is tanked by the agentic purchaasing.

That's amazing, frankly.

I wrote a simple bash script the other day to handle a video encoding queue, with this line:

if [[ $(date +%s -r "$file") -lt $(date +%s --date="1 min ago") ]]

It's running on Debian 12 but to imagine that if it were running on Ubuntu it would have failed?

Wild that this wasn't caught as soon as the dud utility shipped in a distro. I would have expected somebody's scripts to have failed, they ran it under bash -x and thought, "Oh, boy," then off to file a bug.

I like the idea of using Rust and the idea of Software Engineering. But together.

We heard a while back about Google making a nondestructive book scanner that used puffs of air to turn pages and multiple cameras with stitching algorithms.

Is there a home version that people can recommend, product or build plans?

I have at least a hundred out-of-print books, some on taboo subjects, that I'd love to be able to scan and lend out privately.

Frankly this would be a good item to lend around; I'd only need one for a few days a year.

One good example is ModemManager. It can't exit until either the modem is on-line or it times out, generally because you either don't have one or it's not connected to anything. Why it doesn't start out by checking to see if you have a modem and if not exit right away I don't know. Personally, one of the first things I do is disable and mask it so it doesn't even try to start because it's been well over a decade since I last needed it and nuking it that way makes a significant change for the better in the boot time. HTH, HAND.

If Palintir don't think a pricey college education is worth paying for, then I guess they don't want to pay for it. College grads pull higher salaries for those extra years of education, whereas highschool grads can be hired more cheaply.

There's a sizable online sentiment that 'blue-collar' (highschool + on-the-job training) has been unfairly devalued, and in other contexts many people seem to agree that college is largely a waste of time. Yet in the context of Palintir, since it is 'evil,' everybody will adopt the opposite opinion immediately and presume that offering workers a job directly out of highschool is abusive.

To be fair there's a common way to compile Lua to JVM bytecode so it's likely just a Java front-end, not using the basic interpreter.

Back in the day there was a craze to port Lua, Ruby, Perl, Groovy(!), to run as Java front-ends. Not many got put into production outside of Lua.

However the real point here is that it's now "tell me why I shouldn't use Rust" time.

Moving ABI might be a reasonable objection for a small team but Cloudflare has over a hundred engineers on this so it's not a problem.

They get speed and memory safety in exchange for learning "The Rust Way". Seems like a good engineering tradeoff.

IMO Rust is still for the top 20% of engineers so Java's "solid middle" is still quite safe.

Right, "fundamentally" nothing is exactly knowable

https://en.wikipedia.org/wiki/...

I thought that until I learned that they need weekly maintenance tending.

Somebody would need to build an automated battery watering system for homeowners who go away for a long vacation and forget to water their houseplants.

At some point it's too Rube Goldberg to be usable. Now, a few square miles of grid-scale ... somebody could make a business case where land is cheap and sun and water are plentiful.

Don't worry, they are probably getting paid 300b by Oracle, 250b by microsoft, and 38b from Amazon so it all will work out nicely.

A lot of the deals lately seem to be company A and B pay each other X amount of money and pretend that is big revenue despite relatively little net money exchanging hands.