Discussion ***Official*** 2025 Stock Market Thread 💰

[Red Squirrel]

Silver is nuts I wish I had bought more. Bought in around 30-34 CAD it's at around 87 CAD now. What I have is worth almost 10k.

 

[cytg111]

@Indus , tell me you didnt dump rocketlab?

Also, Barrick Mining, you go girl .. wroom wroom.

 

[FelixDeCat]

Market has a tummy ache.

 

[dullard]

Silver is nuts.

Nuts is right. I sold 3.5 kg today. But I don't know at which spot price the check was written. I'll find out next week.

 

[FelixDeCat]

January might be bad on AI unwind fears. If QQQ drops below 570, expect Trump to manipulate in favor of bulls. Juicy dip.

 

[Red Squirrel]

Nuts is right. I sold 3.5 kg today. But I don't know at which spot price the check was written. I'll find out next week.
View attachment 135027

Wow 3.5kg that's going to be a nice chunk of change either way.

 

[FelixDeCat]

I've reading that the spread over traseasuies to lend on AI bonds is widening so Oracle's costs to borrow are going up. Lower tier borrowers pay way more like CoreWeave.

The difference is these are not individuals they can just dilute stockholders heavily if things get ugly. CoreWeave just did some major dilution recently.

 

[FelixDeCat]

We probably turn lower Monday but dip buyers may rush in for trade, but the question is will we go back down to the lows or flat to higher? That is the real question. Seasonally, we should finish flat to higher, but the Oracle news of pushing out data center build out to 2028 may continue to weigh and reign in the AI trade which has the QQQ up 100% in 3 years.

If the Nasdaq 100 closed on Friday, up 21% for the year, that is more than 100% up in 3 years because of the AI bubble. Looking at the chart below, you can see the notorious DOT.COM bubble meltdown, followed by 9-11, which caused negative returns for 3 years. Then you have the 2008 crash of 40%, then the 2022 interest rate moves in response to out-of-control inflation. Other than these 5 times, the QQQ went higher 19 out of 25 years:



So even if we do have a good 25% pullback in 2026, you definitely want to dollar cost average in IMO. 2027 and beyond may be good years and you want to be ready when it does.

 

[IronWing]

I’m reading the financial statements for another Canadian mining junior. According to their home page, they are doing great, bounding towards boundless gold and glory. According to their financials, they are flat ass broke and sinking fast, losing more money every quarter than their total market capitalization. They also have grammatical errors on their home page.

 

[Charmonium]

It's starting to seem as if AI may not be ready for . . . not a very lot.

As I understand it, the secret sauce is data and modeling you use to train the AI. But since large and very much "uncurated" flotsam and jetsam is so seamlessly mixed in - a little like finding a glazed cockroach riding a confectionated maggot ball in your protein shake - well . . . I seem to recall some about garbage in and garbage out.

I have no doubt that AI, when properly and exhaustively trained for fairly technical and sophisticated purposes, does indeed have exiting potential. But that's not where we're at. We're riding over the mescaline river through the pretty, happy . . . trees? To Baba Yaga's house.

 

[dullard]

It's starting to seem as if AI may not be ready for . . . not a very lot.

I basically agree. AI is going to be massively useful. Just at the moment, it isn't there.

Companies are massively investing, racing to be the #1 in AI. The thinking is probably that whomever gets the market first will be the winner. But, that is often not the case. Many times the first players create a market but fail to capitalize on it (at least not enough to justify the massive spending rush to create the market). The second company to swoop in and get the market often is the winner. Apple is a great example. They are usually second to the market. But they do it so well that people love it and believe that Apple was first.

As for no one really using AI, this article has been making the rounds for the last week: https://www.extremetech.com/computi...-goals-because-almost-nobody-is-using-copilot

 

[FelixDeCat]

CoreWeave.

No margins, tons of debt, run away costs. Dilution is probably the only solution.

The market waking up to the nonsense of datacenter "profits". CoreWeave 5 day chart not pretty.

 

[FelixDeCat]

Looks like Santa Claus has fled the country and took the rally with him. Very rare December sell off on AI bubble fears IMO.

Tonight Micron is expected to report blow out numbers but someone has to pay those outrageous prices, and the data center build out will add this headwind to its costs as well.

Meanwhile CoreWeave down again on various concerns.

 

[IronWing]

My community is doing our part. The county board just approved a new data center for which we have no water or electricity.

 

[FelixDeCat]

And I heard some idiot just say on CNBC that while DC demand is twice the avaliable power grid of Texas, electricity prices will come down because more power will lower costs.

They are not building renewable but mostly natural gas burning generators...because Earth has an unlimited supply of water, cement, gas, etc.

Water is already in short supply and some billionaires are hogging that too as there is no law against it.

 

[GodisanAtheist]

My community is doing our part. The county board just approved a new data center for which we have no water or electricity.

- What about your water and electricity? The data center could always use that...

 

[FelixDeCat]

- What about your water and electricity? The data center could always use that...

The inevitable...the new data center job will be to donate your lifeforce for 12 hours a day as a human battery 🔋

No guarantees you get unplugged at the end of your shift because someone did not show up to replace you and some billionaire needs to finish binge watching Mad Men.

 

[cytg111]

Nah, not like this?

Remember the movie Elysium? With Matt Damon?

Yea well, you gotta start somewhere

https://futurism.com/future-society/tech-billionaires-city-startups

 

[dullard]

Looks like Santa Claus has fled the country and took the rally with him. Very rare December sell off on AI bubble fears IMO.

Just in time to be told tonight that the economy has never been better! A++++++++

 

[biostud]

My community is doing our part. The county board just approved a new data center for which we have no water or electricity.

It will be the Indian kind of data center

 

[Charmonium]

I basically agree. AI is going to be massively useful. Just at the moment, it isn't there.

Companies are massively investing, racing to be the #1 in AI. The thinking is probably that whomever gets the market first will be the winner. But, that is often not the case. Many times the first players create a market but fail to capitalize on it (at least not enough to justify the massive spending rush to create the market). The second company to swoop in and get the market often is the winner. Apple is a great example. They are usually second to the market. But they do it so well that people love it and believe that Apple was first.

As for no one really using AI, this article has been making the rounds for the last week: https://www.extremetech.com/computi...-goals-because-almost-nobody-is-using-copilot

Personally, I think that AI - as in the actual software as opposed to training methods and content - is only half of the picture. But as far as I can tell, that aspect of the process tends to be proprietary and therefore, non-standardized.

One of the things I learned recently is that there is a lot of competition regarding that training data. Here's what I got from searching "inputs for AI training."

Inputs for AI training are primarily massive, diverse datasets (text, images, audio, video, numbers) that are cleaned, structured, and labeled (or sometimes unlabeled for unsupervised learning) to teach models patterns, with human-crafted examples (prompts and ideal responses) guiding fine-tuning for specific tasks like summarization or classification, ensuring the AI learns useful, aligned behaviors.

 

[FelixDeCat]

I use free online AI photo editing tools. They are impressive...sometimes. But Im not paying for that. I also don't GPT.

I use Alexa plus but its annoying sometimes.

 

[FelixDeCat]

They are probably going to keep pumping on autopilot even with questionable CPI because that is what seasonable algos do.

Who knew a president could just declare program after program after program with one pen and the market just gobbbles it up.

 

[cytg111]

I hoping MindMedicine pops next year ... with a stroke of a pen

 

[dasherHampton]

I read somewhere that someone actually gave Micron a $500 price target . I that happens I might be the guy at the Lambo dealership.

I've always dreamed of owning a supercar, at least for a short period of time. The biggest problem seems to be is that those cars have a pretty low clearance. Hit a pothole - $50,000 repair.

That scares me. But - nice option to have.