The Stock Market thread

[Arles]

The stock market has a down two weeks and now Nvidia is Cisco, AI investment is now sub prime lending in 2008 and will need a bailout.

Time will tell, but all this makes me chuckle.

[RainMaker]

I think the issue is these companies are highly leveraged and doing some sketchy accounting. Meanwhile no one knows how to make money from AI. At some point you have to show the world how you plan to make money from this technology that hasn't lived up to the hype.

And it doesn't leave much confidence when they are all hinting at needing bailouts (and even got one earlier this week).

[GrantDawg]

Originally Posted by RainMaker

I think the issue is these companies are highly leveraged and doing some sketchy accounting. Meanwhile no one knows how to make money from AI. At some point you have to show the world how you plan to make money from this technology that hasn't lived up to the hype.

And it doesn't leave much confidence when they are all hinting at needing bailouts (and even got one earlier this week).

I don't think it is a question if there is money to made on AI. As Arles points out there's already money being made in certain sectors. There will be other sectors as well that will eventually be able to integrate it into their business. The question is as you point out. Will they be able to make the amount of money to overcome the massive outlay they are putting into it. That's where we get into the bubble. Just like the dot com bubble. Their were businesses that survived and made huge money on the web, but there were also huge numbers that lost money when the bubble burst. That is what is likely to happen here. A few companies are come out as survivors and billions of dollars are going to be lost in the companies that don't.

[RainMaker]

Who is making money from AI besides Nvidia? Every single major tech company is reporting massive losses in AI. There has been no actual plan for how to make money from what exists.

[bhlloy]

It’s a race to AGI. The first large company to get there becomes the big bad corporation in all those dystopian movies. And everyone is betting their future they can get there first.

[RainMaker]

That is not happening anytime soon. Even the AI companies have backed off those predictions.

[Arles]

I just said my day job has already massive increased our defect detection with AI and saved two full positions. We just started last year and have made about 10x on our investment.

If you can't save money with AI in manufacturing, shipping, warehousing, advertising or customer service... well, you aren't using it right. Amazon and Google are going to get a ton of ROI on their AI investments and that just plays right into the pocketbooks of NVidia and AMD.

You have to believe one of the following:
1. It's a shell game and companies like Amazon, Google, Oracle, ServiceNow, Microsoft, etc will all be taking massive losses, Nvidia will go bankrupt with no orders and AI investments will be the death of this economy as all these companies were idiots.
Or
2. These companies have a plan to leverage this technology into saving money and revenue growth. Some will succeed a ton, some others will fail like every new technology over the past 20 years.

It's crazy so many people think #1 is the answer. But the cool thing is, time will tell. It's not going to be a smooth ride, but it will end up with a bunch of winners and my money is on Nvidia, AMD, Microsoft, Google and Amazon.

[flere-imsaho]

Originally Posted by RainMaker

I think the issue is these companies are highly leveraged and doing some sketchy accounting.

This is absolutely the issue. Whether people can make money off of AI or not (evidence suggests they can), the issue above is exactly the problem that led us to the multiple financial problems of the 00s, including the eventual (and frankly inevitable) Great Recession.

Unfortunately neither party has ever seemed interested in policing the balance sheets of "too big to fail" organizations, so I guess we're in for another round of this bullshit.

[RainMaker]

If things are so great with AI, why are they asking for handouts?

[Fidatelo]

Losing money on AI right now means nothing. Amazon famously lost money for years before ever turning a profit. The key right now is to become one of the 2 or 3 survivors that entrench themselves in all the businesses that come to rely on their products. At that point it doesn't matter if the AI gets much better than what we currently have, they can just start slowly charging more until they are in the green.

[RainMaker]

Amazon had a plan for how to make money. There is no plan to make money from AI at the moment. OoenAI is projected to lose half a trillion dollars through 2030. How are they making that back as their expenses continue to rise?

[PilotMan]

Originally Posted by Fidatelo

Losing money on AI right now means nothing. Amazon famously lost money for years before ever turning a profit. The key right now is to become one of the 2 or 3 survivors that entrench themselves in all the businesses that come to rely on their products. At that point it doesn't matter if the AI gets much better than what we currently have, they can just start slowly charging more until they are in the green.

What do you think Elon wanted all that government data for? And why?


Seems obvious that its value, like his the billions he lost on Twitter, netted him access to something followed by the tangible payoff, that will far exceed the value of that investment. And it will firmly entrench his company in that race and he will leverage his own Ai in a way that ensures himself a seat of control, outside the actual government, for the rest of his life.

[Fidatelo]

Originally Posted by RainMaker

Amazon had a plan for how to make money. There is no plan to make money from AI at the moment. OoenAI is projected to lose half a trillion dollars through 2030. How are they making that back as their expenses continue to rise?

I don't think OpenAI specifically will come out of this, but look at something like Gemini or any number of the coding tools like Cursor. Right now they are cheap for companies to use and likely lose money. And yeah, they are not perfect tools by any means. But they do add _some_ value, and maybe more importantly, companies are being sold that they add even more than they do. So those companies get everyone a license for Cursor or Gemini or whatever and realise some small gains, then lay off a few people or just stop hiring juniors. Fast-forward 5 years and you have a workforce that is totally reliant on these tools, and that's when the providers just start turning the screws. $20/mth? Now it's $30. Then $40. Then $100. All of this is still cheaper than trying to rehire or train people, especially since half the staff who were somewhat competent without these tools have already left or been let go.

[RainMaker]

Originally Posted by Fidatelo

I don't think OpenAI specifically will come out of this, but look at something like Gemini or any number of the coding tools like Cursor. Right now they are cheap for companies to use and likely lose money. And yeah, they are not perfect tools by any means. But they do add _some_ value, and maybe more importantly, companies are being sold that they add even more than they do. So those companies get everyone a license for Cursor or Gemini or whatever and realise some small gains, then lay off a few people or just stop hiring juniors. Fast-forward 5 years and you have a workforce that is totally reliant on these tools, and that's when the providers just start turning the screws. $20/mth? Now it's $30. Then $40. Then $100. All of this is still cheaper than trying to rehire or train people, especially since half the staff who were somewhat competent without these tools have already left or been let go.

They already charge and they lose considerable money per paid user. And yes they could raise prices, but their costs are going up dramatically too as more complex tasks requires more computing power and energy. So it's not about charging $30/month to make it profitable, it's charging hundreds a month for consumer users to break even (and even that might not be enough).

It has value to people and companies. LLMs are really cool technology when you step back. But it's not how you achieve AGI which is so far away that it's silly even talking about it. And consumers are not exactly big fans of the technology either. You're going to need to see some major advances that we just aren't seeing for it to be anywhere near what they promised.

Then you have energy costs. Where are we getting the energy and clean water to expand this at the scale that is needed? Who is investing trillions in energy? Where is the water coming from?

And finally, there are a ton of legal issues still. As it stands, LLMs are plagiarism machines. They are stealing other people's works. How are they going to compensate people and artists for that? You are talking trillions in copyright violations. Who is responsible if an AI gives deadly medical advice? Or an AI system commits a crime? There's a lot of liability issues that need to be sorted for something that just doesn't make money or provide much tangible benefit to society.

[Fidatelo]

Originally Posted by RainMaker

They already charge and they lose considerable money per paid user. And yes they could raise prices, but their costs are going up dramatically too as more complex tasks requires more computing power and energy. So it's not about charging $30/month to make it profitable, it's charging hundreds a month for consumer users to break even (and even that might not be enough).

I'm not sure where you are getting your numbers from but I find it difficult to believe that no company will be able to combine price increases with efficiency gains to ever find a profit point from LLMs. All of them? No. A couple of them? I think so.

Originally Posted by RainMaker

It has value to people and companies. LLMs are really cool technology when you step back. But it's not how you achieve AGI which is so far away that it's silly even talking about it. And consumers are not exactly big fans of the technology either. You're going to need to see some major advances that we just aren't seeing for it to be anywhere near what they promised.

I don't think either of us has a clue how close or far we are from AGI. Also, that's not really what I'm discussing here anyways. I believe there is a path to profitability from LLMs in basically their current (functional) state.

Originally Posted by RainMaker

Then you have energy costs. Where are we getting the energy and clean water to expand this at the scale that is needed? Who is investing trillions in energy? Where is the water coming from?

Water and power? Don't worry, there is a world of weaker nations to exploit for those ends. This is the oil of the future. Your country will just invade mine to get the clean water it needs.

Originally Posted by RainMaker

And finally, there are a ton of legal issues still. As it stands, LLMs are plagiarism machines. They are stealing other people's works. How are they going to compensate people and artists for that? You are talking trillions in copyright violations. Who is responsible if an AI gives deadly medical advice? Or an AI system commits a crime? There's a lot of liability issues that need to be sorted for something that just doesn't make money or provide much tangible benefit to society.

And finally laws? Hahahaahaa! Laws! Good one. To paraphrase another great AI, Allen Iverson:

"I mean, listen, we're talking about laws, not a profit, not a profit, not a profit, we talking about laws. Not a profit. Not, not … Not the profit that I go out there and kill for and work every day like it's your last. Not the profit, but we're talking about laws, man. I mean, how silly is that? … And we talking about laws."

[albionmoonlight]

I do think that the Ai/Techbros are making a financial mistake by hitching their wagons so completely to one side of the political spectrum. Other big industries (retail, banking, etc.) keep friendly with both GOP and Dem politicians.

But the AI bros have all these really weird ideas about remaking society in the image of very divorced dudes who think that colleges shouldn't teach humanities classes. So there really isn't a home for them in the Democratic party.

If the Dems do end up in power again and that happens to be when the AI companies need their bailout and/or getting an exception to environmental and IP law, they might not find a very receptive audience.

[GrantDawg]

Originally Posted by albionmoonlight

I do think that the Ai/Techbros are making a financial mistake by hitching their wagons so completely to one side of the political spectrum. Other big industries (retail, banking, etc.) keep friendly with both GOP and Dem politicians.

But the AI bros have all these really weird ideas about remaking society in the image of very divorced dudes who think that colleges shouldn't teach humanities classes. So there really isn't a home for them in the Democratic party.

If the Dems do end up in power again and that happens to be when the AI companies need their bailout and/or getting an exception to environmental and IP law, they might not find a very receptive audience.

Sire they will. Democrats are whores. They will take their money and do their bidding.

[sovereignstar v2]

Longtime consumer brand Crucial doesn’t want to sell you RAM anymore when they can sell to AI companies instead. Hope you like your RAM expensive!

[Arles]

Yeah, there's no way I would buy DDR5 RAM right now with the cost. DDR3/DDR4 are still somewhat reasonable. I would bet some discounted pre-built machines will go for way less than the current cost of their parts.

[dubb93]

Originally Posted by Arles

Yeah, there's no way I would buy DDR5 RAM right now with the cost. DDR3/DDR4 are still somewhat reasonable. I would bet some discounted pre-built machines will go for way less than the current cost of their parts.

It's insane. September 2024 I bought an AMD x3D processor in a bundle from Micro Center with a motherboard and 32gb of DDR5 included for less than you can buy 32gb of DDR5 today.

[RainMaker]

I bought this for $189 in July.

Are you a human?

[Arles]

Disney investing $1 billion into AI to allow characters on Sora AI video generator.

[albionmoonlight]

Awesome growth numbers out today.

Maybe this will get the Fed to start thinking about inflation again instead of just the “cUt no MattEr WhaT!” mantra they’ve had recently.

[Lathum]

Originally Posted by Arles

Disney investing $1 billion into AI to allow characters on Sora AI video generator.

This is nuts considering in the past Disney has sued daycares that painted murals of Mickey on their walls.

We went to Disney several years back and did their Halloween party. They very clearly state in the TOC that you may dress like a Disney character but absolutely are not allowed to pose for pictures with other guests. My wife went as Joy from Inside Out and really pulled it off. Several people asked for pictures and we had to turn them down.

Disney is INSANE about their intellectual property. My guess is they did the calculation they aren't going to be able to prevent it, so may as well have a say in it.

[flere-imsaho]

Originally Posted by albionmoonlight

Awesome growth numbers out today.

Do we believe numbers being produced by the Trump Administration? As I understand it, the most recent inflation numbers, usually based on a grouping of 30 factors, contained just three this time around: Used cars, new cars, and gasoline.

[NobodyHere]

SP500 down 140 right now.

Thanks Trump. Hope pissing off Europe over Greenland is worth it.

Or is this Biden's economy today?

[RainMaker]

Don't think those cheap mortgages are coming folks.

https://www.cnbc.com/2026/01/20/us-1...iff-trade.html

[flere-imsaho]

Once again we'll have to elect a Democrat in 2028 to fix the economy that the Republicans have ruined.

[albionmoonlight]

[flere-imsaho]

Well, that's cheery, albion. :P

[NobodyHere]

Nice to see the SP500 up two percent today. Hopefully Trump doesn't say anything stupid and ruin it.

[Lathum]

Originally Posted by NobodyHere

Nice to see the SP500 up two percent today. Hopefully Trump doesn't say anything stupid and ruin it.

It is all a mirage and designed for his rich buddies to make more money. It will tank again once Iran calls his bluff.

[Young Drachma]

Originally Posted by Lathum

It is all a mirage and designed for his rich buddies to make more money. It will tank again once Iran calls his bluff.

grifts all the way down

[JPhillips]

This will work out well, I'm sure.

The FCC has banned the sale of all consumer-grade routers made outside the US. There is a waiver process, and the goal is to build them here.

[Mota]

Time to reconnect your AOL accounts, that's probably all that a "Made in the USA" router could handle.

[NobodyHere]

Can we stop with all the winning yet?

[albionmoonlight]

Originally Posted by NobodyHere

Can we stop with all the winning yet?

Sadly, I don't think we're close to done.

[flere-imsaho]

DOW 40,000, here we come!

[JPhillips]

Oil's going up, though, so everything evens out.

[NobodyHere]

Another winning day on Wall Street. Thanks Trump!

[Arles]

Just remember this time last year (4/2/25), the S&P was down 20% since late February due to the tariff scare. It was up 37% the rest of the year. It's been down around 10% since late Feb this year and it could go down another 4-5% still, but buying now will probably pay off by the end of the year.

If you look at the history of international "crisis" on the market the past 6-7 years (Covid, Ukraine, Tariffs, now Iran), buying when the fear was at a high level has always payed off.

[cuervo72]

Buy low, sell high? GENIUS! I wonder why everybody doesn't simply do that.

[JPhillips]

Over time that's definitely true. In five years the market will be up compared to today. But everything I'm reading about oil and related shortages coming in April makes me think we aren't at the bottom yet.

[Danny]

Originally Posted by Arles

Just remember this time last year (4/2/25), the S&P was down 20% since late February due to the tariff scare. It was up 37% the rest of the year. It's been down around 10% since late Feb this year and it could go down another 4-5% still, but buying now will probably pay off by the end of the year.

If you look at the history of international "crisis" on the market the past 6-7 years (Covid, Ukraine, Tariffs, now Iran), buying when the fear was at a high level has always payed off.

Youre likely correct though I hope things stay low until after the mid terms and the rebound comes after. We at least need some greater checks for Trumps last two years.

[RainMaker]

I think a protracted war with Iran is about as bad as it can get for the economy and I'm not sure how the country can get out of it at this point. The only hope for it going up is figuring that out.

[NobodyHere]

Today I be walking around like:

[JPhillips]

Everything is vibes. Only two non-Iranian ships made it through the Strait before Iran closed it again and yet oil dropped by 50 dollars a barrel.

[Ksyrup]

That's what is so crazy to me. It feels like the collective wants any excuse to drive the market to new highs and Trump and his administration is doing any and everything they can to strangle the economy.

[PilotMan]

The funny thing is when Biden was in charge, if there was even a dog fart in the middle east, fox was banging the drum about it being a total melt down and the markets 'dropped on uncertainty'. It really is crazy how disconnected it feels to the world right now.

[Ghost Econ]

This world is so stupid. They're pivoting from something people need to something no one wants and the people who own it has their stock price go up 700% for announcing something pointless. We are literally creating "money" out of thin air and pretending everything is going fine.

Stock Chart Icon

Struggling shoe retailer Allbirds makes bizarre pivot from shoes to AI, stock explodes more than 700%