• Welcome to ASR. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Master AI (Artificial Intelligence) Discussion/News Thread

Well, it's also important to understand that "language" has many different meanings. There's the languages we speak, the languages we use to interact with different things in our daily lives (programming languages, the interactions with our pets and cars ...) etc etc. Narrowing it down to human spoken language is NOT appropriate when it comes to LLMs.
LLMs don’t depend on language as meaning. They depend on language as pattern. To an LLM, English, Python code, legal text, music notation - all look similar. They’re just structured sequences of symbols. "Language" is the interface - there's no real intelligence behind it. "Large" is because of the huge quantity of structured and unstructured data aquired in training; i.e. the model has been exposed and internalized a massive amount of "patterns".

Yes, of course.

Note however that you've quoted the first two sentences from my reply to @Salt who (incorrectly) argued that 'understanding ... genuine language meaning' was a prerequisite for generative LLM output. So 'language' refers to human (written) language in that instance.
 
Last edited:
What memory is it that is expensive? I'm assuming it must be the more recent DDR5. Some DDR4 ram modules I bought in 2018 are still cheaper now than they were then.


EDIT - yes, the DDR5 equivalent of my DDR4 ram is about 55% more expensive.
I am comparing with the prices during summer and fall of last year, and not with the prices at release. The Crucial Pro OC DDR5 64 GB 6000MHz kit was apparently selling at about $210 prior to the RAM supply crisis. I bought it for $533 a couple of weeks ago. By the end of last week, its price had jumped to $679 or so

at Best Buy, from where I bought it
https://www.bestbuy.com/product/cru...-3-0-amd-expo-desktop-memory-black/JX8PSKCQX3

at Amazon
https://www.amazon.com/Crucial-6000MHz-Overclocking-Desktop-Compatible/dp/B0DSQVNBD5?th=1
Amazon seems to have run out of stock of ANY DRAM for which it is listed as the seller. All that's left are third-party sellers.

at Micro Center
https://www.microcenter.com/product...nnel-desktop-memory-kit-cp2k32g60c40u5b-black
The cheapest 64GB DDR5 kits at Micro Center are $500 at the moment. A TeamGroup kit I checked was listed at $500 on Sunday night, but has jumped to $600 when checked Monday night.

The YT channel Paul's Hardware has taken to tracking GPU and RAM prices monthly. The RAM price analysis begins about 6 mins 40 secs into the video below.
 
The Crucial Pro OC DDR5 64 GB 6000MHz kit was apparently selling at about $210 prior to the RAM supply crisis. I bought it for $533 a couple of weeks ago. By the end of last week, its price had jumped to $679 or so
I was wonding if that was related to USA tarrifs - but no, that kit is currently £645 over here. Though that does include a 20% VAT markup. So £516 before tax.
 
Micron Technologies, the marketer of the Crucial consumer DRAM brand, is out of the consumer DRAM market/business to focus on GPT data center customers.:


December 3, 2025 at 11:40 AM EST

Micron Announces Exit from Crucial Consumer Business.


Consumer-level Crucial branded products will no longer be available after the end of February 2026. Stocks are probably already being drawn down with the concomitant increase in prices.
 
Consumer-level Crucial branded products will no longer be available after the end of February 2026. Stocks are probably already being drawn down with the concomitant increase in prices.
If we were watching the trotters, few horses/jockeys to keep an eye on, would be

As of 2025:
Yangtze Memory Technologies (YMTC), which holds 4.5% of the global NAND flash memory market share, making it a key Chinese competitor to global giants like Samsung, SK Hynix, and Micron.
ChangXin Memory Technologies
(CXMT) is the leading Chinese player in the DRAM market. While its current global DRAM market share is around 5%, CXMT has rapidly scaled production capacity—capable of producing the equivalent of 10–13% of global DRAM output based on wafer capacity. Also heavily 'invested' in DDR5 and developing high-bandwidth memory (HBM) for AI applications.
Hua Hong Semiconductor (China): ~2% is the second-largest Chinese foundry, focused on mature nodes
GigaDevice (China): ~3% is a fabless flash memory and MCU developer.

imo: Like what BYD is doing -in the EV market- across the world.
 
The Gamers Nexus YT channel led by Steve Burke does an invaluable deep dive into whether alternative sources of NAND and DRAM, such as CXMT and YMTC, could rescue consumers from AI-infrastructure build-out related supply shortages planned by the big three Micron, Samsung and SK Hynix.

 
I rather believe that the collapse of the Open AI, Nvidia, Oracle and CoralWeave circle jerk will rescue us from over priced DRAM.
Even if some of the current companies collapse financially due to over investment - don't think for one minute that AI is going away.
 
Even if some of the current companies collapse financially due to over investment - don't think for one minute that AI is going away.
I''m not sure where I stated that, I do hope that Microsoft copilot goes away though.

It will be interesting to see which of the current consumer AI systems survive or come out on top.

I don't think it will be Open AI who are bleeding cash and resorting to advertising so may not survive the free product phase of venture capitalism, or Microsoft Copilot who were caught defrauding subscribers to pay for it, is too pervasive throughout their products, is generally disliked and in my experience very poor, and Microsoft seem to be losing focus on their other products with disastrous consequences.

The smart bet in my opinion would be Google Gemini as Google have the experience and training data.
 
Last edited:
The smart bet in my opinion would be Google Gemini as Google have the experience and training data.
Really?
You are going to be betting on one single, demi-god for your entire digital life?

HowManyEggs2PutInBucket.jpg

AIl-knowing and possibly AIl-controlling? :facepalm:
 
So the GPT brains were at Davos recently, along with the rest of the swell, and Mr. Amodei gave them his vision of the GPT-based future:

"Dario Amodei, the CEO of Anthropic, told an audience at Davos that AI models would replace the work of all software developers within a year and would reach “Nobel-level” scientific research in multiple fields within two years. He said 50% of white-collar jobs would disappear within five years."

Well known squint and GPT huckster, OpenAI CEO Sam Altman (who was not at Davos this year), has said; ". . . we are already beginning to slip past human-level AGI toward “superintelligence,” or AI that would be smarter than all humans combined."

Back in July, Altman gave a 'presentation' at the Capital Framework for Large Banks conference, where he told the stuffed-shirt bankers that certain job categories would be completely eliminated by AI advancement.

“Some areas, again, I think just like totally, totally gone,” he said, singling out customer support roles. “That’s a category where I just say, you know what, when you call customer support, you’re on target and AI, and that’s fine.”

How many of those people believed this hype? How many of them looked at the guy on their right, then looked at the guy on their left, and came to the realization that those are the guys that Altman and Amodei are talking about and then, suddenly, they realized that they, too, would be displaced by GPT?

When will the time come that the GPT touts like Altman, Amodei, Musk, Zuckerberg, et. al., walk into a board meeting to be told by the BOD that the board feels that laptop here can run the company better than they can? So, therefore, you're out and laptop is in charge.

gptceo.png


Will these guys ever come to the realization that all the hype they are spewing about jobs being eliminated by their GPT applies to them as well?
 
Really?
You are going to be betting on one single, demi-god for your entire digital life?

View attachment 507529
AIl-knowing and possibly AIl-controlling? :facepalm:
Betting yes, hoping no.

We are it the free drugs stage of the venture capitalist / drug dealer playbook.

Whoever can stay afloat and start charging exorbitant prices to cover what has been invested will win.
 
Betting yes, hoping no.
We are it the free drugs stage of the venture capitalist / drug dealer playbook.
Whoever can stay afloat and start charging exorbitant prices to cover what has been invested will win.
If it was not for the android-bubble some of my cohorts occupy (w/o resistance); I'd be leading a totally google-less digital existence!
Sorry, I tend to overreact upon hearing acquiescence to google.
My bad!:(
 
So the GPT brains were at Davos recently, along with the rest of the swell, and Mr. Amodei gave them his vision of the GPT-based future:

"Dario Amodei, the CEO of Anthropic, told an audience at Davos that AI models would replace the work of all software developers within a year and would reach “Nobel-level” scientific research in multiple fields within two years. He said 50% of white-collar jobs would disappear within five years."

Well known squint and GPT huckster, OpenAI CEO Sam Altman (who was not at Davos this year), has said; ". . . we are already beginning to slip past human-level AGI toward “superintelligence,” or AI that would be smarter than all humans combined."

Back in July, Altman gave a 'presentation' at the Capital Framework for Large Banks conference, where he told the stuffed-shirt bankers that certain job categories would be completely eliminated by AI advancement.

“Some areas, again, I think just like totally, totally gone,” he said, singling out customer support roles. “That’s a category where I just say, you know what, when you call customer support, you’re on target and AI, and that’s fine.”

How many of those people believed this hype? How many of them looked at the guy on their right, then looked at the guy on their left, and came to the realization that those are the guys that Altman and Amodei are talking about and then, suddenly, they realized that they, too, would be displaced by GPT?

When will the time come that the GPT touts like Altman, Amodei, Musk, Zuckerberg, et. al., walk into a board meeting to be told by the BOD that the board feels that laptop here can run the company better than they can? So, therefore, you're out and laptop is in charge.

View attachment 507560

Will these guys ever come to the realization that all the hype they are spewing about jobs being eliminated by their GPT applies to them as well?


I do wonder what all these chaps are thinking - when all the jobs are gone, and no-one has any money, who is going to buy whatever it is that they are selling?
 
Perhaps we will all be just staring at our screens ("god") vacantly.
Waiting.
 
I do wonder what all these chaps are thinking - when all the jobs are gone, and no-one has any money, who is going to buy whatever it is that they are selling?

That part puzzles me also.
 
I do wonder what all these chaps are thinking - when all the jobs are gone, and no-one has any money, who is going to buy whatever it is that they are selling?
I've noticed that more of the 'tech bros' are pushing a glowing future with Universal Basic Income. Though they don't explain how this will come about.

The idea has been mooted for many years by some economic theorists on both left and right. Though the details of how it would work are very different for the two 'sides'. :)

I could quote what Gemini 3 Pro told me when I asked it "I seem to recall that both left and right political bodies have advocated a Universal Basic Income if AI causes mass job losses. What are the different perspectives on this?" - but maybe you can ask similar to your favourite AI?

Personally I can't see a path to the sunny uplands of a glorious future.
 
I do wonder what all these chaps are thinking - when all the jobs are gone, and no-one has any money, who is going to buy whatever it is that they are selling?
Musk has been talking about some sort of 'universal income' for more than a decade, starting with 'universal basic income' then moving to 'universal high income' on the basis that AI would further improve productivity, but without giving any details about how he thinks it will happen. It's not a unique idea though - several countries have run trials in recent years, starting with Finland in 2017. There are links to several of them in this BMJ article which looks at a basic income in terms of physical and mental health benefits during COVID. Wales ran a trial between 2022 and 2025 but the results aren't in yet AFAIK. Again the hope is that it will improve physical and mental health, saving healthcare costs in the NHS which is also state funded. Another trial in England was proposed (Jarrow, Finchley) but I don't know whether it ever started. Notably that trial was proposed under a right wing government (at least by UK standards) looking at the economics - would it save the government money overall?
 
Back
Top Bottom