Elon Musk's xAI wants $1B cash infusion in exchange for equity shares The world's richest man is begging for cash again, this time he'd like $1 billion for his recently formed AI outfit, with individual investors told to offer at least $2 million for a slice of the equity pie. In a filing yesterday with the US Securities and Exchange Commission (SEC), Elon Musk's xAI said it had already nabbed …
He expects other employees to spend at least 40 hours/week in an office/on a worksite, for each of those companies.
There's 168 hours in a single week. Even if we exclude everything after Neuralink, because he basically just revived the idea of the hyperloop and then dumped it on gullible idiots who've never heard of pneumatic tubes. That's still 6 companies. So, 6 companies X 40 hrs/week per company = 240 hours. Now, I admit my math isn't always the greatest, but I feel pretty confident when I say that 240 > 168. That means there's no way he could be living up to his own requirement for other employees of spending 40 hrs/week on the job, and that's even assuming he never needs to eat, sleep, take a shit, or anything else. And of course we already know he spends hours every day shitposting on Xitter. So, he's not even necessarily spending 40hrs/week at any of those businesses actually doing work.
"He expects other employees to spend at least 40 hours/week in an office/on a worksite,"
I believe you demonstrate either a fundamental lack of understanding or simple jealousy here.
One of the basic benefits of owning a company is that a person decides how much to work whilst writing job descriptions that state how much their employees should work and what their wage will be.
If a person wants to have their nose in ten troughs (like many executives and politicians) that's between them (and possibly the boards of the companies*).
Welcome to capitalism 101.
*I exclude voters having any say in what their representatives do as "an employee of the State" because the entire system seems corrupt.
Just expecting that a "leader" would hold themselves to at least the same standard they hold everyone else. Especially an asshole who likes to claim he would sleep on the manufacturing floor to make sure things were ready. I don't think it's in any way unreasonable, let alone makes me jealous, to expect that if the boss requires the underlings to be in the office 40hrs/week, the boss isn't out on the golf course 35hrs/week. The boss should be putting in at least 40hrs/week along with everyone else. Shared sacrifice and all that.
He's not spread too thin, he's just found another make of snakeoil to sell as people have started to work out the rest he's selling doesn't work.
The Boring Company and Hypeloop are doing nothing, Neuralink is being investigated for animal torture, Tesla's car software is dangerous according to a whistleblower, Starship still isn't ready for crewed flight unlike Orion which has crewed flight planned for the next mission, and Musk has turned Xitter into a dumpster fire. Where else is there left to turn except AI?
"Coat
Re: Spread too thin
"Where else is there left to turn except AI?"
Blockchain!!!!"
Umm, Elon is the lead defendant in the largest lawsuit ever over his pump of Doge. Not only is Elon listed, so are all of the companies he operates. It may not go well for his defense if he continues on with crypto at this time.
Orion is not the best comparison. The first time its life support system will be tested in zero gravity is with crew on board. Different safety standards applied to Dragon and Starliner and will apply to Starship. NASA's GAO is now reporting what rocket enthusiasts knew in 2016: the next human Moon landing will be no earlier than 2027. There is plenty of time considering SpaceX runs the only US operational crew rated space vehicle.
One of the things I do not get is commenters feeling the need to give Musk credit for SpaceX. Not 8 hours aerogems pointed out Musk cannot spend 40hours/day at each of his companies. Plenty of others have pointed out that he spends all his time on Twitter spreading hate and promoting scams - the things he is actually good at. SpaceX is run by skilled rocket scientists. I refuse to let that arsehole take credit for their work.
I have always believed the way to win is to attack in force at a point of weakness. SpaceX is raking in profits and will have no serious competitors for at least two years. Twitter's stockpile of cash used to pay interest on its biggest loan has almost run out. Musk is running up debts and fines like he can escape them through bankruptcy. He has been working through the list of things competent CEOs never do because they would allow piercing the corporate veil. That would convert Twitter debts into Elon's debts.
Musk's Tesla shares are already leveraged to the limit so the courts will wrench payment of his Twitter debts from elsewhere. It would take a stable genius to run up so much debt at Twitter that he loses the majority of voting shares at SpaceX. That is the sort of ability I would credit to Musk. If you want to help, look at who advertises on Twitter and tell them their customers are not RWNJs.
He used to listen so he could talk like a rocket scientist. He has a genuine interest so made the effort to understand. I think that made the big difference between SpaceX and Blue Origin. Recently he has spent so much time at Twitter that ringwatchers (people who spend their lives watching live streams from Boca Chica) figure out some of where Starship is going before Musk catches up.
SpaceX has no problem hiring the best in the industry for example Kathy Lueders and Bill Gertenmaier. This is why I have so much confidence that Starship HLS will be safe by the time Artemis III is ready.
"Exactly: SpaceX is down to Gwynne Shotwell, Tom Mueller et al."
Tom has long since left SpaceX and is very critical of Starship. The amount of money that Elon has estimated SpaceX will spend on Starship in 2024 (likely underestimated) is far more than the company brings in from Falcon launches and other business.
“commenters feeling the need to give Musk credit for SpaceX”
You are incorrect, he does deserve credit for SpaceX. He didn’t invent the rocket, but he did understand earlier than anyone else the key metrics and what needed to be done:
#1 The key metric is $/payload. Not ISP, or the rocket equation. OldSpace spent *sixty years*, and upwards of three hundred billion dollars, fannying around trying to find “the best propellant”, or the lightest titanium valve. None of which matters, or is even counter to the real objective. Until Musk, nobody stood back and really grok’d at a gut level that while a 30:1 propellant ratio looks both glorious and wasteful, what’s really wasteful is having a workforce of two thousand people optimise and tick boxes for a five year program, all to save an amount of fuel that would keep a 747 jet flying for just two days.
The best propellant is cheap and available, and with fewest logistics challenges, not the one with highest specific impulse. The best structural material is anything that doesn’t have “aerospace 10x price and machining difficulties” attached to it. The best manufacturing process is one you can just order parts from standard machine-shops down the road, and doesn’t require “specialist aerospace skills” that aren’t hireable.
#2 There’s nothing special about rocket design per se. Plenty of chemical engineering uses supersonic flow of reactants, in really large quantities, of stuff that blows up. The problem with rockets, is that we simply don’t tend to do enough of them, to get over that initial design hump. His solution is just to *fly a lot of prototypes*. If you do a lot of something, it becomes routine, and routine is what makes things reliable. If you pre-analyse to death in simulation, you can reach the same quality and reliability but not only is it slow it’s *fragile*. You’ve *never* flown enough things that you can change a single nut or bolt without it being a nail biter.
#3 The thing that makes launchers seem expensive, is that we throw them away. Every design/parts cost is subsidiary to that. Launchers aren’t actually expensive, that’s a failure of perspective. Ariane 6 is cheaper to build than an Airbus A350, and is much less complex. Plus an A350 uses more than twice as much fuel in a week (look it up). But because we don’t throw them away, A350s pay for themselves in the work they do. And Ariane doesn’t.
#4 Iteration cycle time is key. Every space program I’ve ever been a part of, has a huge marching army problem. Every six months delay costs six months worth of salaries. Cutting the design cycle time to 3 months cuts your costs so heavily you wouldn’t believe. Once the prototype is built, having everyone hang around “to make sure” is a money-sink, to the point where it’s literally better to just launch test and watch a few million dollars burn. Ariane employs 8000 people. At the moment, they aren’t doing much else than developing Ariane 6. Their loaded rate is about €110 per hour (yes, it is). Every month a prototype sits waiting for checklists to run, costs Ariane €160 That’s twice the build cost. Now do you understand just how *crazy* it is to spend six months chewing away on paperwork, rather than just launching the damn thing to test?
It’s the fact that the traditional rocket companies and their funding agencies still haven’t taken those points to heart, makes them continue to fail, in the metric that matters. That’s what Musk got paid billions for. It might not seem much to you, but it’s more than anyone else has done.
Well, let's see, what other sci-fi concepts from the 1950s have we got?
He's already doing "colonies on Mars" (for specific values of "doing", anyway).
How about antigravity, interdimensional travel, warp drives, and time travel? It doesn't matter that they're all based on magical thinking, since Musk appears to be no stranger to that mode of thought.
Most LLMs that are trained on real world data like Twitter have problems where they degenerate into crude, racist, misinformation-spewing abominations. Researchers have struggled quite a bit to overcome those issues.
Musky just labeled that behavior as a feature and shipped it.
This post has been deleted by its author
Be careful what you wish for. As Bogeyman-in-chief, Musko is a fine source of entertainment, providing a never ending source of clickbait news, keeping innumerable journalists in work.
And who would we mock without him? Bezos, Zuck, Gates, Cook? Nah, tame, dull, borrrrriing. The world needs Musk as the pantomime villain of the tech sector, in the same way it needs Fat Boi Kim as the pantomime villain of dictators.
Like any good conman will tell you, why spend your own money on something when you can convince some rich sucker to pay for it?
I won't call AI a fad, but I will call it a solution in desperate search of a problem. Right now no one can really define what AI is, what it's good for, or pretty much anything else about it... but everyone is tripping over themselves to try to add it to their product, whether it makes sense or not. Eventually the actual useful applications for the tech will separate from the chaff, but not before a lot of people will be bilked out of billions, maybe even trillions.
Right now, "AI" is nothing more than a marketing term for pattern-matching and generation software. There's no intelligence, just like "machine learning" isn't really learning anything. I could fill a hard drive with the contents of all the works in the British Library, and do some fancy indexing on them, the hard drive wouldn't have learned anything, any more than the paper in those books has learned something when the words are printed on them (unless you happen to live on the Discworld, in which case, ooook!).
On one hand it should meanwhile be obvious to any investor that his intensifying god-complex, FU-attitude and right-wing conspiracy tendencies can no longer be ignored.
All of those pose a clear danger to anyone who plans to actually earn money by giving some of theirs to Mr Musk.
So it will become interesting to observe if Mr. Musk's ongoing harsh treatment of investors/victims of his X/Twitter personal blog will result in enough people thininking hard about this proposal for his personal "non-woke" AI bot.
On the other hand - as the article seems to indicate - investment decisions driven by greed and rational reasoning usually do not seem to be tightly coupled. Or coupled at all.
Everyone being hurt by this investment should have known better.
I'm out of this and prefer to invest in popcorn...
Years ago there was a friendly community of Tesla owners. They were mostly wealthy, not particularly evil and worshipped Musk. These days you can find Teslas with stickers on expressing disapproval of Musk. It is taking far longer than I anticipated but Musk repeatedly whacking them with a clue bat is having an effect on that community. Perhaps he will say GFY to them as well as former Twitter advertisers.
If you use the cesspit that we call "The Internet" to train a large language model, you will encounter the computing axiom known as "garbage in: garbage out".
I don't personally use X, or its previous incarnation, but given that the size of a post there is limited and given that it has a reputation for hosting extreme ends of the viewpoints held by those able to operate a smartphone app, I'd suspect that, at last, people might stop describing the output of an ML model trained on X's content as being anything like "Artificial Intelligence".
Suggestions for what xAI should, more realistically, be called are invited.
Last time he did "funding secured" it was on Twitter. This time it is an SEC filing and already the SEC felt the need to say they "won't attest to the accuracy of the information in the filing". Perhaps the change in venue is because he actually knows better than to pull the same scam in the same place twice in a row. The other possibility is most of the people with a spare $2M to flush stopped using Twitter.
Musk's idea of "binding and enforceable agreement[s]" probably involves sending all of SpaceX's cryo-coolers to hell.
A poorly defined investment with no measurable milestones or outcomes, from a man who has done such a bang-up job with Twitter "X".
Where do we sign up? Don't all rush at once.
Since we can't even agree on what an AGI would be, or do (and indeed, write down a formal definition of what "intelligence" actually is), how does he plan to go about creating one?
If I had anything to do with the SEC, I'd be watching very closely to what happens to every cent of the money invested in this.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
