They’ve been fantasizing about that ever since “computers” started growing in accessibility - in the 1960s…

Fantasizing wasn’t the best choice of words - I often understate what I mean to communicate at an attempt at humor. I should have said "everyone started fantasizing becoming so obsessed with intelligent “AI” that they’re willing to dump a significant portion of the world’s resources just because… "

The current crop is just the first time such things have been delivered with something resembling “average” human responses.

That’s more or less what I meant by “patterns of language that seem relevant to a given input”. I was attempting to understate this in order to exaggerate the villainous eagerness and stupidity of greedy, rich fucks.

The LLM craze is a natural maturation point of the AI field

I don’t see why that is. Using ML to generate models that accurately perform specific tasks is orders of magnitude away from attempting to feed the entirety of human text into ML and expecting superhuman intelligence to emerge.

now it’s expanded into foundational models (FM) which you would still probably just call LLMs because most people don’t know the differences.

While ML and “AI” is not my field, I’m fairly certain that what I was attempting to describe in layman’s terms in my literal first sentence were these foundational models you are referring to.

FMs are getting close to that point of a magical universal computer that you can tell it to do anything about anything and it just works.

I have no direct experience outside of LLMs and I don’t really take issue with what I understand FMs to be, so long as they keep their scope narrow and focus on accurating completing specific tasks to assist humans. As soon as we hand off control and trust it blindly without extensive trials ensuring it’s reliability and failsafes in place to ensure inaccuracies are caught I start raising concerns.

My only experience is with LLMs - a few, minor attempts to “test the waters” of the major, publicly available LLM models. I’ve been frustrated with my search results and glanced at the AI results. Work gave us Gemini licenses and I used it in similar, desperate situatiuons for coding help and help with Google products foolishly thinking that if any LLM designed to help with such tasks would be passably useful it would be the LLM of the company that owns the products I seek help with. Unless something has changed drastically in the last month or so, every interaction has been a roll of the dice to such an extent that my occasional “testing the waters” caused me to jump out and avoid it as much as possible. I simply can’t trust it to not halucinate and gaslight me.

What I see as the problem is moving way, way, way too quickly in trusting language models to do anything even remotely important. Human communication is extremely nuanced, complicated, fluid, and imperfect. Humans misunderstand each other during communication even when we have the context of in-person visual/audible cues and interpersonal history.

Maybe it’s because I’ve only ever had at most a comfortable income but I truly don’t understand the mentality of needing so much money.

I don’t get paid as much as my peers but I make enough to be comfortable. I am my own department and, aside from emergencies and other high priority situations, I manage myself and choose what to work on when. I have a decent work life balance. Because I make enough to be comfortable (in large part because my landlord promised not to raise our rent - early in the COVID lockdown - if we were “good tenants” and has managed to keep true to her word) I don’t feel the need for more. That balance is worth not making the 20% more a year I might get somewhere else because I can’t guarantee I won’t have a shitty boss that doesn’t let me have that work/life balance.

I was excited about the idea of purpose-built systems trained on specific datasets to be help find complex patterns to diagnose diseases or suggest potential molecules for specific purposes.

Then the LLM shit started and everyone started fantasizing about intelligent “AI” just because it was able to reproduce patterns of language that seem relevant to a given input. Some of those funding it kept chasing that dream and are convinced that, if they just throw more compute at the problem, they can evolve the renaissance AGI that can do anything. Then they can fire every worker and be bazillionaires with robot slaves and never have to work another day of their lives… and fuck everyone and everything else.

It’s amazing what we can ruin when we let greed and selfishness drive our society.

This is something I’m really struggling with. I’m finally at a point where I have sufficient income to start putting it away. I have over 20+ years until I’m of retirement age and I’ve got a state pension but no matching contributions so I’ve never started any personal retirement account. I’ve always felt like it was just another scam for the wealthy to profit off of the working class.

Of course, they get to charge me a fee - take their cut. However, can I be sure that they have my best interests in mind? If they can leverage all of the investors an index fund, or whatever the fuck I’m invested in, and throw us under the bus to benefit themselves or a high-value client that’s actually paying attention and throwing around orders of magnitude more money… will they?

Ignoring the scam angle… what the fuck can I invest in, without putting significant time and energy into, that isn’t diametrically opposed to my ideals? Tech, power, pharma, banking? For lack of a better term, anything making money these days is either Evilcorp or likely to be put out of business by or acquired by Evilcorp. Do I want to profit off of that shit, or even help raise their stock price?

I could always put it in a CD or something at a credit union and know that I making substantially less for my retirement than virtually every other option. Yay.

So… when there is potential for someone to take advantage of something, it should be denied to everyone lest they happen to be the type that might take advantage? That more or less negates all social safety nets, charity, and acts of kindness.

Interesting that there really are employees that flourish and work more effectively when they do so from home.

How much do those who take advantage cost the employer vs those that benefit the employer? What the net gain or loss? Is it impossible to unobtrusively measure this? Maybe those who take advantage can be put on an improvement plan, brought into the office, or terminated without banning the practice.

I have somewhat similar concerns. I’m not as worried about sanitized Linux as I am about new mandates entrenching Microsoft, Apple, and Google as the only valid options. Even it it is an enormous pain in the ass for everyone, including those big three, it would infinitely preferably for them to more widespread adoption of alternatives.

Just propose solutions/mandates that are fundamentally incompatible with GPL and FOSS ideals, or deeply contentious within open source communities and you can do irreparable damage to the growth of Linux and any space that needs to adapt to those new mandates. Linux moving into education? Pretend it is needed to protect the children. Linux moving into government? Pretend it’s needed to protect security or efficiency. Linux moving into the workplace? Pretend it’s needed to protect AI or liability or synergy or whatever the fuck gets CEO dicks hard these days. BAM - Linux gets hit with massive internal strife and splitting of vital communities and resources. I know it was already absurdly contentious before, but seeing what happened when storing a users age hit systemd really worried me.

I think it’s already been kind of ongoing by co-opting or even creating “open source initiatives” from the business world who ultimately just jump in when things look mature and rapidly implement profit extraction and enshittification.

We are not far from general AI

You mean AGI? Yeah, no. I don’t believe you.

My microwave has an option to turn off all the beeps, and it’s a landlord special so I imagine others can do the same. Check your model and see if it has that option.

For mine, I literally need to hold down I think the number 3 for 5 seconds. It’s in no way obvious.

Golden.

Essentially, the employees most excited and inspired by “visionary” corporate jargon may be the least equipped to make effective, practical business decisions for their companies.

“This creates a concerning cycle,” Littrell said. “Employees who are more likely to fall for corporate bullshit may help elevate the types of dysfunctional leaders who are more likely to use it, creating a sort of negative feedback loop. Rather than a ‘rising tide lifting all boats,’ a higher level of corporate BS in an organization acts more like a clogged toilet of inefficiency.”

And yet, they did it, very quickly, and will do so again when the market shifts again.

These aren’t typical conditions. There is a world of difference between “Yo, here is a truckload of money. Give me all your RAM for the next two years!” (AI right now) and “We need to align our capacity to the current needs of the market in order to not flood the market with product we can’t sell.” (normal conditions).

Look at the memory market before this AI shitshow. It’s cyclical as the fabs try to match demand without flooding the market. It typically takes several quarters to ramp capacity up and down, causing shortages and then oversaturating the market because they can’t do it fast enough. Returning to that kind of production is also likely to be more difficult than it would otherwise be to adjust while already in it. There are so many unknowns in this wild situation. How much DRAM should they commit to making?

Anyway, I’m not even a business person so what the fuck do I know. I’m getting tired of this conversation. It would be great to be optimistic, but that’s not something I’m capable of right now. Best of luck. I hope it pops and prices plummet.

Please indicate the point at which we no longer agree.

• The price of consumer hardware, like RAM and SSDs, is high right now.

• Supply for consumer hardware is low right now.

• The reason that the price of consumer hardware is high right now is because supply is low.

• The reason this supply is low is because manufacturing capacity that used to be committed to consumer variants of this hardware has been converted to manufacturing datacenter variants of this hardware.

• The reason for the manufacturing shift is that, because of the AI bubble, demand for datacenter hardware is high.

• It takes time and money to convert manufacturing capacity between consumer hardware to datacenter hardware.

• This conversion also includes changing investment, R&D, and employment priorities. Companies like Micron (one of the three major/noteworthy memory suppliers in the world) have cut/gutted their B2C divisions to focus their resources entirely on the more profitable datacenter B2B efforts.

• Datacenter variants of this hardware are substantially different than consumer variants and are not interchangeable.

• Higher prices, slimmer margins, and lower sales have caused some consumer electronics businesses to cut back, focus on narrower market segments, pivot to other markets, or exit the market entirely.

If we agree on all of the above, I’m not sure why you are confused.

If the bubble pops today, it very well may cause demand for datacenter hardware to crater, assuming they don’t get creative and find a way to pivot hard toward other SaaS somehow. Since the parts are not at all compatible with consumer electronics, this would not provide an instant increase in consumer hardware supply nor would it crater consumer hardware demand.

If the crash is bad enough, some of the businesses that pivoted to B2B/datacenter hardware may fail or have to be bailed out by their governments. Even if they all survived, fewer consumer product businesses exist in the market to sell to consumers and they have fewer product lines in development. Companies like Micron, who abandoned their established brands and/or relationships with resellers or partners that would use their parts in consumer products, will need to rebuild those things. Manufacturing would need to be converted, updated to the latest consumer hardware needs, taking more time and money.

It would take quite a while to rebuild consumer hardware supply. Without that supply, demand will not be met and prices will be high.

My apologies! I didn’t realize you were the arbiter of what I do and don’t need. I feel so relieved now that I can just ignore all of the demands on my life and just hand over such authority to some opinionated jackass I met on lemmy.

Why would it be immediate? Where is the new supply coming from to alleviate the demand?

I feel like you didn’t read what I wrote. If the bubble pops today, how long do you think it will take for prices to drop?

…why would it need to be?

It doesn’t need to, but if it were that would be the only reason I’d say prices might drop anytime soon. A glut of used consumer-compatible parts would push prices down. That or maybe if the rising Chinese suppliers manage to ramp up and find a way to enter the western market.

The price is currently high and is rising because resources and manufacturing capacity are limited. Those who own the capacity have found that providing for a small number of companies that are flush with cash and will throw money around just to ensure their competitors don’t gain an advantage is far more lucrative than providing for consumers or businesses that integrate parts into consumer devices. The entire market segment is shifting away from consumer and focusing on datacenter hardware.

The longer this goes on, the further the major players will be from being able to pivot back to consumer products… and there are only major players in the memory and NAND industry. You can’t just form a new memory or NAND company and start manufacturing this stuff. It takes years and a lot of investment to build the facilities and the kind of capacity we’re used to.

Edit: I’ve also seen a number of non-tech folks excited for cheap used datacenter memory and gpus to flood the market after the bubble pops, as if the parts were at all compatible with consumer devices. I wanted to make sure that was not part of your calculation.

Once the bubble pops, assuming it doesn’t take economies with it, none of the product will be compatible with consumer devices. Manufacturing will have to be reoriented back to consumer products, then those parts will need to be manufactured, then the rush of people trying to get the parts will have to pass. THEN maybe prices will come down.

I suspect the datacenters will just pivot and repurposed to rent consumers “cloud compute” and cloud subscription services and continue to fuck the entire consumer market for years to come.

But then again I now hate everything so maybe I’m just pessimistic.

Just in time for RAM, SSD, and HDD prices to skyrocket and make personal computers unaffordable.

I guess if you can afford one now, at least you’ll be able to repair it.

Allegedly an incorrect rumor.

Windows Central

Fucking LOL. https://www.idmerit.com/blog/idmerits-data-breach-fail-safe-architecture-game-changer/