Hacker News Reader

Just keep in mind this is all funny money !

The valuation of those 2 companies combined should be < $10 billions if we are very very generous.

X is heavily in debt, diminishing user base, and bleeding money. 0 profits

xAI is yet to make any actual revenues outside of ... X. Also bleeding money

Although I would admit, that these companies are not valued based by their economics, but rather on political power they provide to the owner.

I’m guessing Tesla is paying xAI to train their FSD model?

That would be a shocking conflict of interest - a public company paying a private company, which is owned by the CEO of the public one ?

All while the same CEO claims that the public company is the leader in the field ?

Even by Tesla & Musk lawlessness standard that would be too wild.

Don't worry, the Tesla board is totally on top of things.

We already have multiple news stories about xAI receiving 10,000s of GPUs that were originally meant for Tesla. That's about as close to "public company giving private company owned by CEO money" as you could possibly get without cartoon sacks of cash.

> Even by Tesla & Musk lawlessness standard that would be too wild.

Not to be overly snarky here, but, uh: what standard? They have proven over and over that they could not give less of a shit about either societal norms or the laws of the nation.

Now that you say that, Tesla is already paying the Boring Company, to build a highly advertised ... short tunnel for Tesla Giga factory.

Which they spend 3 years or so building and which seem completely useless. The only big unknown would be how much Tesla did Boring Company for that.

So yeah...

The entire stack is based on TSLA valuation, which is going to tank.

IMO this is going to be meteoric in nature. As in the meteor hits the ground.

If the world wasn't perceptibly irrational, I'd short.

Yep, I also believe that we will have a Minsky Moment on the entire Muskonomy.

Given his political involvement, I would assume that the maximum end date for this house of cards is the date when Trump is no longer in power - which is less than 4 years away. The real date is probably much sooner, but who knows when ?

For anyone else who wasn't familiar with a Minsky Moment

https://www.investopedia.com/terms/m/minskymoment.asp

But that then means that Musk has a strong incentive to keep Trump (or some replacement that will similarly let him do whatever he wants) in power past these 4 years.

Buying time to get the rocket finished that he will ride escaping his creditors to Mars. That would be a badly written plot even by the standard of cheesy comic books, but somehow it all makes sense from that angle. (no, I don't believe that he would ever get within explosion range of a rocket engine ignition)

People do not last in Trump's orbit. I'm shocked Musk has lasted as long as he has. I though he'd practically be a Scaramucci.

Every six months is a 50% chance of ejection.

A few hundred billion will buy you a lot of politicians in this climate.

There is also the "first buddy" benefit. The whole thing is being actively rebuilt at a time when Musk has been heading a department which, amongst other things, openly backdoored the treasury. It's pretty uncharted waters.

Stock market graveyards are full of people that shorted TSLA. I remember reading about people who lost it all because of a TSLA short at least since 2020. The lesson here is that markets can remain irrational longer than we can remain solvent.

I remember all the other major shorts.

In those cases, it usually revolved around Tesla releasing either the Model 3 or Y, and they were betting against Musk's execution in general. What wasn't really up for debate (in my mind) was the technological lead or the demand in the marketplace for the car.

Essentially, Tesla was primed in both situations for a huge market they had to themselves: electrified transport.

Now?

- demand is tanking, and the brand is permanently damaged, well, as long as Musk is at the helm.

- competitors are everywhere

- technological moat (with respect to EV drivetrain/construction) is gone. I could argue their cylindrical cells are legacy/semi-obsolete in fact, since IMO pouch cells are superior for LFP/sodium ion, and who knows what solid state will optimally package.

Anyway, things are VERY different in my opinion.

> The entire stack is based on TSLA valuation, which is going to tank.

But it's not is it. Try looking at the trends long term, not just YTD. It's remarkably stable, trending upwards. Yes it gain a bunch of value from October to January/February, which is has lost. Investors seems to have traded on the "First Buddy" effect, that didn't actually do anything for Tesla, so that value has now been removed and we're back on the original trend line.

How the hell drop in sales and lose of brand value hasn't caused to stock to absolutely tank is beyond me. Is someone artificially keeping the price stable? Do investors see something I don't, or they do just not care?

That's what Tesla shareholders get for not giving Musk the $50B that he so much "deserved" to be fully "invested" in the company.

No, Tesla has their own seperate training supercomputer called Dojo that it optimized for computer vision. Its one of the top 20 supercomputers in the world by compute.

That’s all marketing. They have had lots of issues getting it put together and working, and is more akin to really really really self driving in 2018 we were told.

BTW: find it: https://top500.org/lists/top500/list/2024/11/

Yes - it's really really close to FSD as you say ! Here is also a nice chart they published 2 years ago:

https://electrek.co/2023/06/21/tesla-dojo-supercomputer-is-f...

Definitely on the FSD timeline right now - they are maybe at 1% of what they announced they would be today. But hey, it's Optimus now !

As best I can tell the deal includes paying off $12 billion of Twitter debt. So that's real money, though that seems like it's basically all of the cash that xAI has raised, so if that's the case that's pretty wild.

Edit: I misread, Musk said it's an all-stock deal, so you're right, it's all funny money. I think that means the combined entity is taking on that debt. It's still great for Twitter and bad for xAI investors because the combined entity now has both the $12 billion debt and all the VC xAI raised.

Definitely great for the Twitter investors to be able to convert. I think they took a haircut relative to the 42B that they came in on, as the 45B includes the 12B debt. (42 gets reduced to 33, so 21-22% haircut) xAI also raised 6B in December last year at a 45B valuation, and then in February, reporting was that xAI was trying to raise 10B more at the 75B valuation... so this is where the frothiness of AI helps to mask the fundamentals. Can gets kicked down the road.

Definitely, and a lot of that haircut actually happened in between when the deal was signed and when it closed, too. That was right when interest rates went up and Musk tried to wriggle out of the purchase agreement. So it's a great outcome for Twitter investors, assuming it sticks.

And assuming the xAI stock they traded their X stock for is worth anything.

Also - the wording on the deal: xAI being valued at 80B X being valued at 33B Is the xAI 80B number inclusive of the 33B (45B including debt)? That would back into xAI's valuation moving backwards since the December round. Usually when venture-backed companies tout a valuation after a capital raise, it's post-money... and they're not chest-thumping a 113B number (which they surely would have given 9 figures)

And.. looks like there was an update on the valuation from Bloomberg: Aggregate over 100B, but less than 113B due to some mingled ownership.

https://archive.is/esoEN

The are getting xAI stock instead of cash though and it's not really clear how liquid it will be.

Every time I see deals like this that are going to be such raw deals for the majority of those who invested it feels like a keen reminder that all the talk of market efficiency the pro caps in Washington talk about are full of it

This is a terrible and inefficient use of money. It’s inevitable that this paper house will fall over

Markets are plenty efficient. They’ve just lit billions of dollars on fire that used to belong to people stupid enough to believe in Elon Musk.

That isn’t efficient by any commonly understood definition of the word

Tesla is still worth something like a forward PE of 90 and a trailing PE of 130, among plummeting sales and market share.

It's still wildly overvalued, and at it's unclear how long the market will stay irrational.

It's so hard to make predictions of Teslas value in today's political system! Musk is very very close to power, and it's impossible to say how that will play out. Will Tesla be the only car company without 25% tariffs? Will the government spend billions to buy Teslas? Or will Tesla die and Musk live of space-X?

He's not "close to power". He's cutting your nation into shreds, unopposed.

Megacorp market valuations are extremely bizarre because they're in a sense self-fulfilling. If your company has a market cap in the hundreds of billions of dollars then it has lower capital costs than competitors. If you want to build another factory, you're paying less in interest. Which means you can undercut the competition on price, or expand into grid storage or data centers etc.

Obviously you then have to actually do those things before the value crashes, so whether it works or not depends on whether investors continue to expect you to.

The current value fluctuations have more to do with politics than the company. The media is writing a lot of negative stories right now, but they also have the attention span of a goldfish and ultimately some new Current Thing will replace the existing one.

Ahhh, I dunno, musk is pretty good at fucking his own companies with Nazi salutes and wreckless incompetent destruction. Doesn't really take any big bad media to see what a fucknut musk is.

The media is the thing that characterizes it as "Nazi salutes" and "wreckless incompetent destruction".

The guy is an eccentric with autism, not an angel or the devil. He's not always right. But everything is so polarized right now that people are barely willing to concede that it's possible for someone to be right some of the time.

Teslas have the positive characteristics that they don't run on petroleum and have good safety ratings. They have the negative characteristics that they have privacy issues and you can't get parts or documentation to fix them yourself. A strong heuristic to tell if someone is using motivated reasoning is if the thing they're complaining about is an unrelated political hot button well-known to produce outrage (e.g. allegations of somehow simultaneously secret and open fascist ideology) as opposed to a way they're actually screwing someone illegitimately but in a way that doesn't cause most people to experience an immediate visceral response (e.g. mass surveillance, right to repair).

You can also tell it's that when the target is actually doing a bad thing, but the people objecting only care about the target and not anyone else doing the same thing. This is extremely common with corruption allegations. All corruption is bad, you need systems in place to constrain government officials from bilking the taxpayer, but this is a process problem where you need stronger checks and balances to prevent the government from transferring money to cronies, not a "we need to somehow elect only infallible humans" problem where what people actually mean is that they just want their guy to get in so the money goes to their cronies instead.

Cool story buddy. But musk is still a Nazi salute throwing fuckwad fool. I'm not buying the smirking excuses. He also happens to control a company producing government subsidized garbage pail cars. And he lies about their self-driving capability... over and over and over again. Mainly because he was too big of a fool to recognize that multimodal sensors are a good thing so he's about half a decade behind.

I prefer cars that don't get recalled for glued on pieces falling off and CEOs that didn't pay a quarter billion dollars to buy a presidency.

Plenty of other electric cars being produced by the other manufacturers.

Consider why it feels important to you to convince people to hate someone.

I fail to see the efficiency part. Care to elaborate?

Garbage collection - although not pretty - is still necessary.

I smell margin call desperation. Musk got visibly worried with tanking Tesla stock and pulled out some crazy stops to keep Tesla afloat.

Remember, these megabillionaires are due to stock valuations, and live rich lifestyles on the basis of lending against the value of stock. If that value drops and a margin call forces actual selling, then Musk loses shares in Tesla.

I have to think this stunt is related to Musk needing some liquidity somehow, or being unable to cover private liquidity crunch with public assets.

TSLA is ludicrously overvalued. P/E is 150, based on Q4 earnings which were 1/3 "real" revenue, 1/3 mark-to-market-BTC (aka one shot), 1/3 CAFE/subsidies which Trump will nuke in 6 months.

But the "real" earnings that were already flat in Q4 are seeing a huge drop in EU/CN and who knows in the US. If "real" revenue drops 20%, the "real" earnings might drop 70% or more.

Anyway, that 150 p/e is ACTUALLY 450 in my mind, but after Q1 reports it will ACTUALLY be almost 1000 or worse.

The only thing keeping the stock afloat is AI hype, but 1000 p/e ratios are for people with market dominance and fundamental leads on competitors. Tesla is arguably middle of the road in both self driving and robots.

Tesla is still worth more than it was 6 months ago. I don't think a margin call based on borrowing a few years ago is likely.

When is the first earnings call after everyone realized Musk was a Nazi?

Q1 sales call.

I was trying to stay apolitical but a CEO doing Nazi salute and not being fired is uncharted waters, at least in the last 80-85 years.

I keep going back to the horsemeat episodes on mad men, and the papa johns CEO getting booted. This is simply unprecedented in brand management. Brands take decades to build, and are especially important for car companies.

People tend to only look at the drop from January until now, but if you look back a little further it seems clear that the market was expecting some payout from Musks relationship to Trump. When that's didn't really amounted to anything, the stock price just dropped down to it's September 2024 level and continued from there, as if October - January didn't happen.

What surprises me is that the drop in sales numbers seems to have zero effect on the stock price, at least not yet.

The earnings call for Q1 2025 hasn't happened yet. Those usually happen at the end of the first month of the subsequent quarter.

If it’s an all stock acquisition did it generate any liquidity?

Sort of: it put Twitter's LBO debt on X.ai's books. I strongly suspect Elon had been lending X money to cover interest payments. Both X.ai and X are privately held, so specifics will be hard to come by.

I was thinking margin calls too: https://news.ycombinator.com/item?id=43423201#43425801

He had a sizable chunk of collateralized borrowings against his Tesla shares even before the Twitter acquisition. There was the risk of a margin call during year following the acquisition, but AI became the latest hotness so he was able to spin up a company to cook up some more market bubble.

This doesn’t strike me as an objective assessment of the state of affairs.

You could make the same “bleeding money” comment about many historical social media acquisitions – at least Twitter has advertising and subscription revenue. And it’s similarly suspect to argue that xAI is failing due to lack of revenue, without mentioning the astronomical valuations of its peers in the AI space, many of which are not only lacking revenue but also have no distribution channel comparable to the scale of Twitter/X.

xAI has one of the leading models in terms of investment and user base, and it’s rapidly improving on its evaluation metrics.

Twitter/X is one of the Top 10, if not Top 5, social media properties worldwide. It has hundreds of millions of monthly active users and maintains the lead in public consciousness as the “go to” place for 1-N broadcasting of information. While Bluesky shows promise, it won’t be dethroning Twitter any time soon. When has Bluesky been the source of breaking news?

And you cannot separate these two businesses. Twitter/X and X.ai are fundamentally coupled to one another. Twitter is not only a data source for X.ai but also a distribution and retention channel. There is an “ask Grok” button under every Tweet. And it’s not the only LLM that can be summoned via Tweet – with similar automations from companies like Perplexity proving there is value in the channel – but it is the only LLM with a native integration into the interface and first dibs on freshly-generated user content and breaking news.

You're missing the fact that Twitter has essentially 0 advertising revenue compared to the other major social networks, as all major ad networks have stopped doing business with Twitter (this in turn being because of the very high likelihood that their ads would be shown next to content that their clients didn't want to be associated with, such as images of Hitler).

> diminishing user base

Where are you getting this from? All numbers I see indicate the user base has been growing globally. https://backlinko.com/twitter-users

Mmh, that source is showing a 10% drop in users for September 2023 the last point, which was notably before alternatives like bluesky and others were starting off. Also note that since Musk took the company private we don't really know how accurate these numbers are.

So I'd argue that the statement theat Twitter/X is bleeding users can't be conclusively decided, but it seems your statement that Twitter has been growing globally is incorrect at least for the last reported quarter.

Elon made a post about user growth being stagnant.

And how many of those are real people and not ever increasing numbers of bots?

Elon saying what the numbers are is the most unreliable thing.

Wait, wasn't there a "sweetener" for recent X debt buyers that they'd get XAI stock?

The levels of .. self-fertilizing transactions here boggles the mind. I guess that's the idea.

> and xAI doesn't have much of any revenue or assets.

xAI does have some pretty massive datacenters with 200K+ GPUs.

Fair enough, edited to just revenue. But those are just datacenters bought with VC money - my intent was to say that xAI's resources are basically just the investment they've received.

You can’t just pull out the corporate Amex and buy a world class datacenter. There is considerable value add there.

They used Tesla’s Amex for that one.

We don't know that it's a world class data center, we only know how many GPUs they bought. For all we know, half of them could be staying on shelvez, and the whole data center could run for 5 hours a day: there's no public info on how well they're doing.

Bro why are you so negative on every comment haha you really leak your negativity on musk companies.

It is a fact that grok3 is a strong competitive model. Surely it had to be trained somewhere? They went from 0 to what they have now in 1 year

Doesn't mean the company is worth $40 billion though... Or even a quarter of that.

xAI is a private company. I very much doubt that the current VC investors will ever litigate this. They are all waiting for some other favour for taking this on either explicitly or implicitly. Also, I would like to see a Venn diagram of the investors in xAI and SpaceX.

They're gonna go public and the market cap will be at ~$300B within a year or two