treis

Can you give an example of your searches?

I'm with the GP. To me LLMs are just better search engines. In the most literal sense. They have their own index and can generate links if you want them.

Okay but I didn't say they didn't conclude that. I said:

>no evidence of burial and the evidence (mussel shells) that humans were involved with the site at all is thin

>Layer 1 was the only unit containing identifiable cultural materials and was interpreted as a midden because it appeared to have been formed, at least partially, by anthropic additions of mussel shells. All in situ dingo bones were confined to Layer 1. There were no observable cut or fill features that indicated a hole had been excavated to emplace the Garli

>One of the Garli’s caudal vertebrae, recovered directly below the eroding burial, was sampled for AMS radiocarbon dating and returned an age range of 963–916 cal BP (ANU-75935). A freshwater mollusc shell sampled from the central burial block (UNSW-3415) returned an age of 1,178–957 cal BP. Three additional mussel fragments (cf. Velesunio sp.) were dated from Layer 1 of Zone 1 (UNSW-3414), the Garli burial core’s interface with Layer 1 (UNSW-3416), and from sediment within the burial core adjacent to the Garli skeleton itself (SANU-75235). These produced ages several hundred years younger than the Garli

This seems highly suspect. The paper says there's no evidence of burial and the evidence (mussel shells) that humans were involved with the site at all is thin. That they came from a "feeding ritual" seems to be based on nothing more than speculation.

Seems a lot more plausible that it washed up on a river bank during a flood. Or it died next to a river which subsequently changed course.

A fab costs a billion dollars (really a lot more) and 5 years. It doesn't do anything for anyone today.

I doubt it's less actual throughput in most cases. In a place like Atlanta there's no place where it's bus after bus. The BRT line they built nearby is a bus every 10 minutes. Which being very generous to the bus usage is equivalent to like 5 cars a minute.

It has $25 billion on AI cap expenditure in the S1. So generally looks like a solid deal for SpaceX.

There is an equally impressive ability of abusers to rewrite history to put themselves in the role of the victim. That comes along with justifying all sorts of behavior. The classic "if you didn't make me so angry I wouldn't hit you" is logic that adds up to them.

Google, Apple, Facebook, Amazon, Nvidia, Cisco, Oracle, et al make a bajillion dollars a year. The faangs are the most profitable companies in the world (swap Nvidia for Netflix)

I'd call that completely wrong. It's become something of an urban legend but the reality is that AWS was net new. It was not share Amazon infrastructure with the world

This gets the story of AWS wrong and it should be fundamentally suspect because of that.

It's big because it may take a big swath of people who will actually pay for LLMs out of the market. But for the average consumer they're going to primarily use their phone/tablet and we're far away from that being possible.

Even if it were possible the LLMs are such a gold mine of user data. It's really hard to see that opportunity be passed up.

I don't think it really adds up. It's an inconvenient install location and roofs are replaced every 10-20ish years. It makes sense from an efficiency standpoint but the capital costs outweigh that.

IMHO a pergola or carport is going to be better. You lose solar efficiency but gain the benefit of something that provides shade. Especially as solar panels have become an economical roofing option if you don't care about perfect waterproofing.

These businesses didn't stop growing. They grew by how you say. Steady growth into adjacent markets and acquisitions. Not through innovative new products.

It's just a different mindset. It's no longer move fast and be first. It's well we'll get there eventually. Let's do it cost constrained and with minimal risk.

Ah yes a typical Tuesday for me

They don't pay a fraction more. Levels.fyi says I'd about double my salary and if they downlevel about 50% more.

Think this is the FAANGs entering their stodgy middle life phase like Oracle et al did decades ago. Except for Google they haven't done anything innovative in like a decade. It's time to wring money out of the monopoly and buy companies to add to that.

They show them now but yeah they're another ad driven service in waiting. That's ultimately the end for Meta. When ad dollars go from websites to LLMs.

I dunno I think AI is pretty cataclysmic for the ad supported web. For one, why go to any webpage instead of reading the LLM summary. For two, why look at an ad riddled page when an LLM can clean it up for you.

"That's what the money is for" - Don Draper

Yes. Nothing fundamentally has changed about the US/Europe relationship for 80 years. Trump being a royal prick doesn't mean his message is fundamentally different than what any other president would send. If the US elects Kamala, for example, things quiet down and go back to normal almost immediately.

The first tractors in 1911 or whatever probably did. 50 years on and it was many times that.

Your data shows a 20% improvement. That's $20-100k a year depending on how much devs are paid.

Replace Iran with Iraq and this was all true 25 years ago. And that was an actual failure.

Fundamentally nothing has changed about the world or the relationship between the US and its allies. Once Trump is replaced by someone closer to European social values and less of an asshole the temperature will change. Just like it did from Bush to Obama.

It's definitely not. It's a fundamental shift on how we interact with computers.

It's a tractors on farms kind of moment.

If we suddenly went from rail travel to jets that's exactly what would happen. We'd go from 0 to all the business flights that happen today. Everyone would be under enormous pressure to not be a laggard.

I disagree that this is well written. It's a long and meandering where unchallengeable anecdotes build up to sweeping narrative that doesn't add up.

Iran has gone from a peak of power and it's proxies pulled off the biggest attack against Israel in decades. Three years later most of the leadership involved is dead, their power is at a nadir, and Israel has re-established itself as the dominant power.

That Iran isn't totally incapable of fighting back and hasn't capitulated isn't much of a feather in their cap.

Technology wise the West has a string of victories. SpaceX, AI, Waymo, and Apple are leaps and bounds ahead of any Chinese competitors.

Nothing has changed. The US honest to goodness lost a war 50 years ago and continued to dominate. Not to mention Iraq/Afghanistan. Iran being something less than a perfect and clean victory doesn't fundamentally change anything.

I think this conflates symptoms and causes. You'd still be a code producing cow with a craft brew in your hand. $200 fun budgets, IME, doesn't help that and more often is another sort of obligation.

AI has definitely improved the industrial capacity of the US

At this point Gitlab investors haven't gotten any money.