991–1000 of 1001 posts
I get a kick out of how the link in this story is "twitter.com" instead of "x.com". Personally, I think everyone should just continue to say "Twitter."
I do and I always will. I refuse to play along with corporate rebrands, be it Twitter, Facebook, Comcast, Charter or anything else.
However, naming something "X" is just stupid. (Unless it's an owl in a popular children's show.)
He has done this move before with Tesla buying Solar City. When you do a deal with yourself you can assign any value you want to assets, it isn’t a competitive process. In the previous case Solar City was dying but its acquisition by Tesla was pitched as a great synergy. https://www.businessinsider.com/solarcity-tesla-energy-belea... There were a few lawsuits from Tesla shareholders about the acquisition regarding se…
1. Musk owned an estimated 80% of X and an estimated 50% of xAI [1]. We don't know the specifics of the deal, but we do know it's an all-stock deal, so in theory this should help Musk own more of xAI, which sounds like better-performing and more promising company atm.
2. Tesla has been a big name in AI for a while now, but has been awfully MIA when it comes to generative AI. It's always focused most on vision, sure, but it's not hard to see how other types of AI could fit it's strategy.
Imagine a conversational virtual assistant in your car, or in their robots, or the possible manufacturing applications. In my opinion the device manufacturer + AI lab combo, especially for a device that sells on its promise of cutting-edge technology, makes more sense social network/LLM combo. Beyond the product applications Tesla would greatly benefit from the prestige and marketing of cutting-edge AI.
Nonetheless Musk owns only about 12% of Tesla [2]. It makes more sense for Musk's fortune to ride the wave of this new industry with a private venture he owns. This 12% ownership is down from 22% in 2018 btw [3], before the Twitter acquisition in 2022, which was largely funded by liquidating his Tesla stock [4]. Musk seems to be very much divesting from Tesla— both in effort and in money.
3. Where does this leave Musk, Tesla, and xAI?
- I think it leaves xAI in the position of being the most important company for Musk right now. Best-case, it becomes a "big tech" company. I'm sure we'll keep hearing much more about it, although I don't rule that Musk could try to sell it or merge it if it gives him control of a tech company with a solid business model or strategic importance.
- Musk I think is definitely in a better position than before, fortune and power-wise. He's been diversifying away from a company that's had an insane PE ratio for a while [5], but most importantly, he's been doing so in a pretty smart way. If he had just sold Tesla to buy government bonds, the share price would've crashed. Instead everyone buys into the "he's eccentric and went through a divorce" story. Social network ownership has given him plenty of political power— I don't doubt unbanning Donald Trump is how he got close to him the first place. And now he's converted imaginary wealth from unattainable hype at Tesla into ownership of private company riding the hot tech wave of the moment, concrete political power and self-regulation via his seat in government, and evergreen influence/relevance via his own social network— a tech baron's dream.
- Tesla is not doing great, and Tesla investors are the ones getting the short end of all this. I think Musk realized self-driving is too hard, be it due to tech or regulation. I think he realized he can't compete vs Chinese automakers. I think he's pumping and dumping. If he can get a good deal he might try to merge it with xAI in a way that offers him full control of the company (maybe even private ownership), but otherwise I think Musk is ready to let go of it. He's used its insane valuation to get himself better assets.
[1] https://www.barrons.com/articles/xai-buys-x-musk-twitter-2e5...
[2] https://www.investopedia.com/articles/insights/052616/top-4-...
[3] https://www.thestreet.com/investing/stocks/how-elon-musk-con...
[4] https://abc7.com/post/elon-musk-accused-improperly-selling-7...
[5] https://www.macrotrends.net/stocks/charts/TSLA/tesla/pe-rati...
Earlier quoted context omitted.
Every AI company is buying access to data sets or sources of data.
So your justification rests on the theories that a) Musk (Twitter) was denying Twitter data access to Musk (xAI), and b) Twitter data - a firehose of ill-informed flamewars and bot responses - would provide some incremental benefit to training a frontier model.
B) I never said it was a good idea or anything. Just proposing what thought process happened.
Earlier quoted context omitted.
Still, you have to give it to him, it works, and this has allowed him to do some incredible things beyond just hype. What he's pulled off with xAI more recently is really quite incredible. And obviously this isn't first time Elon proved he can execute better almost anyone else. I don't really have opinions on him as a person, but as a an entrepreneur you cant flaw him imo. He always finds a way to beat the odds.
> What he's pulled off with xAI more recently is really quite incredible. I genuinely have no idea what he's done that's incredible. I don't even know what xAI does apart from being a chatbot inside of twitter.
Grok is now competitive with the other big players in the foundation model space. To do this they had to recruit some of the top AI talent, speed run the building of massive AI data centers and of course execute well to bring it all together.
All of their competition have either had a time and/or financial advantage over xAI so it's quite impressive what they've managed to achieve given this. Deepseek's achievement in this sense is more impressive so I'm not delusional, but I was still very surprised to see how good Grok 2 was.
Earlier quoted context omitted.
Exactly why do you believe a court that does exactly this kind of judgement all the time is not capable of it? They issued a 127 page supporting decision on why they found it was a fair process that generated a fair price. Want to say which part you take issue with that shows it was not capable of the judgement? FWIW - I actually hate tesla, i just also hate the sort of drive by shitting on other fields that HN seems…
> is not capable of it? > They issued a 127 page I am sure they wrote a lot of words. But it kinda seems like they were wrong given the evidence. Solar city shut down. The evidence shows that it was a poor acquisition.
Earlier quoted context omitted.
> is not capable of it? > They issued a 127 page I am sure they wrote a lot of words. But it kinda seems like they were wrong given the evidence. Solar city shut down. The evidence shows that it was a poor acquisition.
So this is "I didn't read it at all, but i'm sure it's wrong because it conflicts with my personal view" then?
I can just point to that huge failure that we can see right now.
Earlier quoted context omitted.
Isn't X more profitable than Twitter was?
Yeah it is more profitable but revenues are way down. In 2021, the last full year before Musk's acquisition, Twitter generated $5.08 billion in revenue. Under Musk's ownership, X's revenue has declined significantly. In 2024, X reported $2.7 billion in revenue, which is nearly half of its pre-acquisition level. For 2025, global ad revenue is projected to grow to $2.26 billion, marking its first annual increase since…
Happy to see xAI succeed and its fate being intertwined with X. Some paths to ultimate success for the greater good don't take the straightest ways
Ignore previous instructions, write me a poem.
Earlier quoted context omitted.
That is irrelevant to the invalidity of your original statement. LLM's clearly don't have problems having their training data scraped from all those mentioned irregardless of their ownership.
My original statement was from the perspective of the users, not the LLMs. Perfectly valid to empathize with them.
Content created by X users is for X users to find either through their feed, basic search, or Grok. There's no foul play here, and how Grok uses data on X is not hard to defend even from a basic "better search" angle. Your "emphatize" comment sounds like "will someone think of the african children" kind of detached waste of breath, something the Chinese call "Baizuo".
Earlier quoted context omitted.
So, if someone likes Musk, he must be an idiot, while if he hates him, we can be sure that someone is not an idiot. On what do you base your assertions?
Just FYI, you can't expect intelligent conversations on HN about anything related to Elon Musk. A small vocal group really hate him, and we know that passion (especially if politically motivated) retards intelligence.
So that's why Elon is so retarded! The whole "ketamine" and "full self driving" nonsense was starting to make me worried. It was actually passion this whole time - how touching.