Artificial intelligence is not replacing human intuition in maths and physics, but reimagining how questions are asked, explored and understood go.nature.com/4aESYuz

Yeah, just 150 billion confiscated dollars will do precisely what is stated here, eliminate poverty and turn the world into the utopia. This is exactly what was missing and now we got the chance. Don't blow it.

This is a careful critique of limitations of LLM-based agents. It is far from clear what solutions exit. Giving them more examples for how to apply high level lessons can help. I doubt there's just some magic mechanism we could bake in.

No. The landing will be hard.

No. The private sector does a better job. Always. Though Bezos has the better overall solution, even if it takes longer.

Probably all parts of mind exist in AI. The problem is that a human and even a fly have a lot of circuits and stake in its own existence that AI utterly lacks.

If you start trusting in gradient, it is only downhill from there.

Good luck.

All cool, but decades too early. The tech is not there. Only bluster, optics, and financial engineering. Hard landing ahead.

Nobody can discover anything with old info. Humans either. Everything that we know was done very, very painstakingly, over a very long time. AI is for now an augmenter. And is getting very good very fast.

Anything you do on Earth you have to add more margin when you do in space. for cooling, robustness, handling vibrations, radiation, micrometeorite damage, can't be serviced, very expensive to launch, high latency. A datacenter in desert with a pipe of sea water way cheaper.

POV: Investors watching their SpaceX stock after buying the largest loss-making IPO in history

Remember when Elon Musk told investors that Twitter would 5x its revenue to $28 billion by 2028 and more than 4x its users to nearly 1 billion? Well, that's no where near happening. Instead he's made even loftier predictions for a much larger company that will go public tomm. We took a look at the skeptics' views of the SpaceX IPO.

It is worse. The SpaceX and Tesla dreams (which are often intertwined and supposed to reinforce each other) are pipe dream. Datacenters in space are way too expensive. Launching stuff in space and space internet is not worth trillions. Humanoids won't be ready for 10-20 years.

The problem is that Elon's schemes will take longer to realize than he thinks. Likely by a decade, if ever. Datacenters in space are not economical no matter what. So wild house of cards, regardless of what other predictions said.

Sure, don't protect the banks. Go bankrupt, see your economy collapse. The Greeks did the right thing. Brought their house in order. Cut the red tape and gov hand outs. Doing better than ever.

SpaceX is the most overhyped IPO of the decade and it will end exactly the way every overhyped IPO ends. Facebook IPO’d at $38 and traded under that for 15 months. Uber IPO’d at $45 and is still below that adjusted seven years later for a while. WeWork tried at $47 billion and ended at zero. Robinhood IPO’d at $38, hit $85, then $7. Coinbase IPO’d at $381 and was at $40 two years later. Rivian IPO’d at a $100 billion valuation with no meaningful revenue and gave back 90%. Beyond Meat. Peloton. Lyft. DoorDash. Bird. Each one a “generational company” the day it priced. Each one a wealth destruction event for retail within 18 months. The pattern is not a coincidence. Hype IPOs are designed to transfer wealth from the people buying the story to the people who built the story. The bankers get paid. The early employees get out. The VCs get a markup they can show their LPs. The retail investor gets the bag. SpaceX is a great company. That has nothing to do with whether it’s a great stock at IPO. Greatness was already priced in five funding rounds ago. You are not getting in early. You are buying the exit. The only IPO worth chasing is the one nobody is talking about. Those don’t exist anymore because every IPO is marketed like a movie release. So the answer is: don’t chase. Wait two years. Buy it down 70% when the lockup unwinds and the narrative breaks. Or don’t buy it at all and put the money somewhere the bankers haven’t already extracted the alpha. Hype is not an asset class. It’s a tax.

Yeah - sure guys - Elon is totally going to load up his $20,000,000 dollar untested AI space satellite onto his Starship rocket that blows up 41.7% of the time. Instead of just building it on the ground for 1/10th the cost and zero risk. Go give him your entire life savings.

The problem is that Falcon-9 was well-designed. StarShip is fragile. Tesla is good electric car. FSD is bad and late (as unsupervised), good at level 2. xAI failed at AI, is now renting useless hardware. It is not all or nothing. Datacenters in space are too expensive.