rlt

The discovering itself doesn’t seem like the interesting part. If the discovery wasn’t in the training data then it’s a sign AI can produce novel scientific research / experiments.

The coolest thing about this is apparently it’s written by Jack Dorsey, billionaire founder of Twitter/Square. https://x.com/jack/status/1941989435962212728

Cool to see he still gets his hands dirty in code.

Lightning network depends on… the internet… so if both clients are on the internet why not just send messages over that?

What makes you think that?

Ok, but that’s a trade off anyone who works for a private company makes, and it’s never called an “effective cliff”.

React was released in like 2013. We were well on our way to the web being unusable with JavaScript disabled by then.

Andrew Bosworth (CTO of Meta) being commissioned as a Lieutenant Colonel in a new unit of the US Army Reserve is even more interesting in this light.

https://x.com/boztank/status/1933512877140316628?s=46

They don’t have voting power.

Massive difference. You can vest and move on, even if you don’t have liquidity, which most private companies don’t for employees anyway.

Meta “will have a 49% stake in the artificial intelligence startup, but will not have any voting power”

Wouldn’t Scale’s board/execs still have a fiduciary duty to existing shareholders, not just Meta?

Previous articles about this suggested the connections were routed through a government datacenter nearby. Is that not the case?

They’re in a pissing match. The relationship between SpaceX and the federal government is mutually beneficial. I would be surprised if either follows through on their threats.

By my estimate, SpaceX has saved the US government 10s of billions by reducing launch costs with Falcon 9, developing Dragon, Starlink, etc.

Can you elaborate?

“Fine, we’ll keep the non-profit, but we’re going to extract the fuck out of the for-profit”

Quite the arc from the original organization.

It seems like the obvious reason they weren’t rescued earlier was the cost to taxpayers was unnecessary when the astronauts were relatively safe waiting, but if that were the case why wasn’t NASA more transparent with their rationale? It leaves room for doubt that it was at least in part politically motivated.

OpenAI doesn’t have the pre-existing business, relationships, domain knowledge, etc to just throw AGI at every possible use case. They will sell AGI for some fraction of what an equivalent human behind a computer screen would cost.

“AGI” is also an under-specified term. It will start (maybe is already there) equivalent to, say, a human in an overseas call center, but over time improve to the equivalent of a Fortune 500 CEO or Nobel prize winner.

“ASI”, on the other hand, will just recreate entire businesses from scratch.

Is putting RAM on the same chip as processing economical?

I would have assumed you’d want to save the best process/node for processing, and could use a less expensive processes for RAM.

One angle is they’ve been responsible for a number of major wildfires in California, so they should be reinvesting any profits in upgrades/maintenance/technology to prevent that in the future.

> Starlink connections in Ukraine are geofenced

This was debunked. Starlink was never active in Crimea because Crimea was sanctioned by the US. He said if the US government asked him to activate it he would. They did not.

> Do we know that guy is bright?

One of the other guys won the Vesuvius Challenge, so I’d say probably.

There it is, the real reason you’re dunking on Starlink, you don’t like Elon.

25ms is fine for the vast majority of people except maybe hardcore gamers.

3ms at the speed of light is 500 miles. Much of what your accessing will be physically further than that anyway.

I have to wonder how much the Chinese government was aware of what DeepSeek was going to publish, and how much they will allow Chinese labs to publish in the future.

Also can’t you just combine multiple classification requests into a single prompt?

Some people like humor.

> An LLM sounds smart but it just picks the most likely response from the echoes of a billion human voices.

This is where reasoning models come in. Train models on many logical statements then give them enough time to produce a chain of thoughts that’s indistinguishable from “understanding” and “thinking”.

I’m not sure why this leap is so hard for some people to make.

Why wouldn’t they? Most big tech cos offer products at multiple layers of the stack.

It would be cool if we could at least park it for a few years until we can bring it back down in Starships.

Maybe Bezos would help fund those missions, even if they used Starship…

I assume SLS’s shelf life is limited and we’ll be using Starship end to end within 5 years or so.