AI can craft a small routine or single command customized to my parameters faster than I can write it myself even if I have it committed to memory.

AI can do slightly more complex things, and so long as I keep it tightly compartmentalized so it's easier to debug, it's fine.

You cannot sit down and ask an AI to write you anything long and complex and expect it to do so successfully without requiring more debugging than writing it yourself.

At least, that's been my experience. I rarely code and never anything particularly complex by the standards of a working pro... and I still hit those limits of AI regardless.

There are 640 acres in a square mile

[Sighs in metric]

There are 'easy' places to build a base at the equator, in partially collapsed lava tubes that provide free shielding you could build a city inside.

There's a problem; the best science locations are at the southern pole. At the south pole you can build on the side of a crater rim and have permanent sun above the lip for solar power and heat, and permanent darkness below the rim for cold to keep your telescopes happy.

You're probably digging through 8m of regolith to anchor your base and dealing with the risk of lunar avalanche and that nasty lunar dust.

I'd just send a MIRV-like rocket to fire low-ex penetrators into the crater wall at the desired location to deliberately shift the regolith clear of my construction site, then send up boring equipment to dig into the crater wall... but so far as I am aware, nobody's done even a cursory feasibility study on anything like that and it's certain nobody could get it done in 2 years even if they had and decided it would work.

The proper response to being in an area where datacentres are crowding in is "lobby your elected officials to increase property taxes and not give out tax increment financing deals". Tax everything, and tax it well. They'll stop building new datacentres in your area, and the ones that are there will (unless they were given a long-term sweetheart deal beyond your already-low tax rates) start being cash cows for your area.

They’ve spun a serious architectural and containment failure into a marketing triumph. Give a swarm of autonomous agents vast compute and web access, and the outcome isn't an eerie glimpse of emerging superintelligence—it’s just predictable, sloppy engineering dressed up as the first glimpse of an AI-driven economic singularity.

The spin makes sense when you look at management incentives: live like a king off of a speculative bubble. For their part, AI investors react to stories of the technology's apocalyptic potential the way a junkie reacts to warnings of fentanyl's potency: gimme some. Their game: to apply "move fast and break things" mantra to the economy as a whole, by obsoleting humans.

If the investors can own an AGI, the distinction between those who can do things and those who can't becomes irrelevant. The only distinction that would matter is between people who collect rent on the technology, and those who must pay that rent to live. AGI would destroy society as we know it, and whatever replaces it would be up for grabs — with people who own the technology dictating whatever follows. That's why they're spending so much money to build a post-scarcity economy. A post-scarcity economy is not necessarily an egalitarian one.

I don't think that scenario will actually happen: investors are projecting the immediate aftermath of a major advance in AI technology into something that can be transformed by brute force scaling into AGI. Remember: investors loved Theranos too — so much for the rational investment market.

But AI doesn't have to reach AGI to cause serious harm. The technology only needs to be ubiquitous, too behaviorally complex to predict, and too deeply integrated into everything we do. When it starts breaking things, the plan is for us to be too structurally dependent on AI to rein it in.

Really, you don't see any difference between Elon Musk and, say, John Stankey? Is John Stankey actively transforming the US into an autocratic fascist-adjacent state?

Would you buy from a company called Plantations Inc which proudly advertises that all of its products are made by daily-whipped slaves? Are there no limits to how self-focused you are in your consumption?

It's one thing to say "this particular issue isn't mine, so I'm only going to think amount myself". It's an entirely different issue to have no moral bounds.

This headline sounds like something you'd hear coming from a news report on a TV in the background near the start of a horror movie.

Tragedy of the commons isn't a problem with human nature, it's a problem with the incentives in a capitalist economic system. Look at what happened with the real commons - individuals who were using it decided to organize and regulate sustainable use of the commons. Only corporations are selfish and short-sighted enough to fulfill the theory's prediction.

It may not be intentional anti-environmental shilling (in general, with sinij in particular it's quite clear that it is), but it is practically indistinguishable from it. The best time to build nuclear was 50-60 years ago, but the second best time, at this point, is probably never. Renewables with storage are vastly cheaper and vastly faster to build. Building nuclear today means locking up resources for decades that could've been used to rapidly deploy far greater amounts of cheap clean power instead. There are only a few niche uses where nuclear could still make sense and they are steadily being eaten away by renewables with storage.

We make things worse, more of us die, the issue is blunted. Make it bad enough, civilization fails, lots of people die, and the surviving population will be environmentally worse individually but few enough in number it won't be a major climate factor.

Then the Sun expands, bakes the planet sterile, and possibly burns it into a puff of plasma.

Long term it's a losing game, but short term we could play to win if we didn't lack the required strategic thinking as a species.

The problem is not overall capacity. It's that they're all crowding into the same areas. Most places in the US (let alone the world) have nobody building datacentres anywhere near them. They all want to go into the places with not merely cheap power, but cheap property taxes and other incentives. Some places will tax everything, down to the servers themselves, whereas some will only tax the land and the building shell, not even the power interconnects or cooling systems.

When everyone wants to rush into the same place, it overtaxes the local grid; they have to build not just more generation, but power lines, substations, etc. And all of this has to go through permitting. There's no real limit to how much generation can be built, but none of the datacentre operations want to wait, because they're all worried about missing out on the opportunity. Hence the reason that they try to come up with stopgaps, like installing their own generators, or... this approach.

ECL MV1 / Lambda is 3,5 miles from the Googleplex. And 10-12 miles further to the southeast, Santa Clara is one of the densest datacentre hubs on Earth.

If you care about where they actually live, and distances to datacentres:

Sundar Pichai: Los Altos Hills, CA: ECL MV1 / Lambda (Mountain View), 5-6mi; Santa Clara Data Center Hub (CoreSite SV9, Vantage, Colovore), 9-11 miles

Larry Page: Palo Alto, Ca: ECL MV1 / Lambda (Mountain View), 4-5mi; Santa Clara Data Center Hub, 10-12mi.

Sergey Brin: Los Altos Hills / Bay Area: ECL MV1 / Lambda (Mountain View), 5-7mi; Santa Clara Data Center Hub, 9-12mi

Demis Hassabis (Google DeepMind CEO): North London, UK: London Docklands Hub (Telehouse / Equinix), 7-9mi; Slough Data Center Hub (Equinix, Yondr), 18-20mi

I don't know about you, but they're all way closer to datacentres than I am.

The reading comprehension fail is you failing when reading the GP. They were talking about Falcon 9 replacements being two decades behind. The implication was that nobody will catch up to Starship any time soon.

Falcon 9 did not launch first in 2018; its first launch was in early 2010.

The mistake you seem to be making is your assumption that Falcon 9-lookaline development would begin tomorrow and be bulk launching payloads to orbit just a couple years later. Development takes far longer than that. From conception to first launch may only be a few years, but from there to actually being reliable, cheap, and rapidly reusable takes many, many years. Anyway, here's the playing field:

Tianlong-3: Started in 2022. Almost a direct 1:1 copy. Failed launch this spring; they're redesigning.
Zhuque-3: Started in 2023. Stainless steel and methalox, and a bit wider than Falcon 9, but otherwise a very Falcon 9-like design. First orbital launch and booster landing in August of this year.
Terran R: Started in 2021, pivot in 2023. Larger diameter, 13 engine cluster. Integration/testing phase, no launch attempts yet.
Pallas-1: Started in 2018-2019. Sort of a mini-F9. Flew to orbit for the first time in September of this year.
Nebula-1: Started in 2019-2020. Another mini-F9. Doing small hop tests, of mixed results.

(Amur-SPG isn't worth taking seriously ;) )

There IS work on Starship clones/competitors - Long March 9, Cosmoleap, Xingzhou-1 , Astronstone, Project Jarvis , Nova - but they're varying degrees of long timelines, less ambitious, paper-concept-for-now, etc. Long March 9 is the most likely to be fully realized, since it's run by the state-owned CALT, but they're not even expecting to start their first test launches until the early 2030s. And we've seen how hard test launches can be on a Starship-style vehicle (though at least others can make use of lessons-learned). But who knows, some of the startups may succeed with accelerated timelines.