The AI revolution is coming

[email protected]

Some of you seriously need to stop living and dying by a daily heat map. We're at levels that we were at months ago. Anybody who bought months ago is doing fine. Learn to zoom out.

[email protected]

The one that raised the standard of living for more people, produces more green energy than the other, the one that has a worker lead democracy, the one that doesn't pretend they have ever been the greatest country, the one that does exclusively produce countries with school shooters...

[email protected]

Chinese ChatGPT outperforms

On par with

on cheaper and less powerful chips as well.

Trained for less money on weaker hardware

[email protected]

Question: Stock rises as a result of its perceived value correct? So if a stock drops below its actual value, and many traders, having the knowledge that it's better to buy the dip when a stock drops below actual value or their peers, buy the dip, isn't it guaranteed that the stock rises?

I don't know how the stock market works, so this might be a dumb question.

[email protected]

As someone who is rather new to the topic: I have a GPU with 16 GB VRAM and only recently installed Ollama. Which size should I use for Deepseek R1?

[email protected]

Not long until those articles are written about the USA.

[email protected]

It’s a good question, but stocks are a lot less logical than you may think. Bubble stocks, especially, are very emotion driven, very much FOMO.

For “boring” stocks, there are statistical models using all sort of financial data about the company, including things like assets, sales trends, new product predictions. Those stock’s prices will appear logical and relatively stable/predictable, but will make significant jumps either way as news comes out that may affect it. Investors buy or sell based on what they predict will happen, and make money if they get it right.

But for “exciting” stocks, this model doesn’t work as well anymore. The stock price is driven more by what people expect it will be in the future. Take Tesla stock as a great example. The price is sky high compared to the value of the company or its sales, very much unlike other car manufacturers what do you do when it’s still a minor company in vehicles sold, but the total stock value is far higher than any other? The CEO has made made some extremely ambitious predictions, that may justify the stock price, but are those predictions really going to happen? You may have heard in the news about investors “shorting” Tesla stock, usually represented as people who don’t believe in the future. The thing is, these were usually investors looking at the actual value of the company, the actual sales, even actual predictions of sales and deciding this does not justify that kind of stock price. They’ve usually been wrong and lost money, because the stock is not fact-driven, but emotion-driven.

What is your perceived value of a stock? What is your prediction of how it will change in the future, based on how well the company is run, what their products, sales, and support are like, how their industry may change, and news that may affect them? For an established, stable stock, you can predict enough to reliably make money, but changes will be small so you won’t make a lot. For a new stock driven by hype, with the potential to revolutionize an industry, predictions are more like a gamble, more based on emotion than fact. However stock price changes in both directions will be big, so investors can make or lose a lot of money very fast, depending on timing.

[email protected]

Makes sense. Thanks for the explanation

[email protected]

The techbro authoritarian society is coming.

https://www.msn.com/en-us/technology/artificial-intelligence/fears-surge-about-panopticon-after-tech-exec-makes-eerie-remark-about-ai-surveillance/ar-AA1xKneg

[email protected]

Okay here we go.

Define intelligence for me please

[email protected]

Run one and ask it about Tiananmen. Post your results here.

[email protected]

I read that in Trump’s voice lol

Lurker

You can try from lowest to bigger. You probably can run biggest too but it will be slow.

[email protected]

We're already there

RFK Jr. Says He’ll Send People Taking Adderall to Labor Camps

Robert F. Kennedy Jr., independent challenger for president and anti-vaccine activist, has a cockamamie solution for drug addicts and people who use medication for depression or ADHD: strip people of their tech gadgets and send them to government-funded "wellness farms" to grow organic food. Doesn't that sound suspiciously... like a labor camp?! Kennedy extolled this […]

Yahoo News (www.yahoo.com)

[email protected]

But it's never a way to normalize that.

[email protected]

100%, AI moves in cycles related to scientific research, same as most other tech innovation funding hype boom/bust cycles.

Incremental scientific research delivers new possible features -> funding hype to be first to market -> engineering builds products based on new tech -> people learn what it can and cannot do through use -> hype dies and funding dries up (you are here for LLMs) -> less funding stops engineering new products and research continues incremental changes until repeat.

It's been that way for AI since the 80s with expert systems, then deep nets, and now LLMs.

[email protected]

Fuck, they want to ban amphetamine-based prescription meds like in shitty authoritarian countries.

P.S. There were more approved drugs (including Adderall which allowed, if I'm not mistaken, allowed only in US) in US than in any other country. About Adderall, I heard that even in EU ritalin and pure (?) amphetamine (in some EU countries only) are allowed.
Also, there are many drugs for multiple sclerosis treatment and they are actively developed in US to be even better. It saves the whole world, I mean, it improves people's lives significantly (however, these medicines are very expensive).

P.S. #2 A few years ago there was Wellbutrin (depression, ADHD, narcolepsy treatment. Innovative new drug) in Russia, and it was registered there officially by pharmacy company GfK, it's manufacturer. Then they shut down production and it stopped being supplied to pharmacies. Then relatives of people who had narcolepsy, lived at that moment in Europe and they offered to bring this drug from Europe to those people in Russia. Then the drug was confiscated at customs and a case was opened against these people, allegedly Wellbutrin contains "drug derivatives", although this is NOTHING written about Wellbutrin, and there are no substances in the composition that could indicate this, there are no such substances officially, even the manufacturer doesn't mention it. It is clear that there is no justice in that country, and that these people could simply be put behind bars. Well, they "scared" people once again and soon released them. It's was just recently. Such biased cases about "narcotic substances" are regularly opened here.

guy

Sure is, but was that a factor for the stock market going down?

[email protected]

I also have 16gb vram and the 32b version runs ok. Anything larger would take too long I think

[email protected]

Incremental scientific research delivers new possible features -> funding hype to be first to market -> engineering builds products based on new tech -> people learn what it can and cannot do through use -> hype dies and funding dries up (you are here for LLMs) -> less funding stops engineering new products and research continues incremental changes until repeat.

TLDR: AI works on goverment funding.