Master AI (Artificial Intelligence) Discussion/News Thread

[JustAnandaDourEyedDude]

I haven't figured out why -when talking about a AI/chatBot- I keep referring to it as a "she".
Could this tendency be construed as some kinky kind of misogyny?

I can imagine many theses and dissertations being written on research into how users think of LLMs or AI Agents. Lots of this research would be funded by AI companies, of course. Does the proportion of users who think of LLMs as a gendered person and prefers a male or a female voice for the LLM to reply with (i'm not talking sex robots here) depend on whether the user's native language is strongly or weakly gendered? How does it depend on the personality of the user?

Will AI Agents be provisioned with a default name that the user can override, or will users always refer to LLMs and Agents as "you"? How many users currently name their multiplicity of agents "Agent Smith"? How will Agents refer to other Agents when speaking with the user? What proportion of users give a name to an AI Agent? Probably more likely to assign names when using multiple agents for different functions, and first name would be easier to remember functionality than referring to them by number. I would guess it would be much more likely for people to name a humanoid robot or animal-resembling robot than an AI Agent.

Will agents be granted degrees by universities if they pass the requisite exams, for a fee natch, and will some users refer to such agents as "Dr." or "Professor" so-and-so? Will users employ honorifics such as "right honorable", "officer", "your highness" or "your excellency" when interacting with specialized AI agents, or just go with "hey you" or "mate"?

Will some people who are dating (other people) judge whether their partner is marriage material by how the partner treats their AI Agents? Will some bosses become less harsh on employees because the bosses have AI Agents they can abuse, assuming the employees have not yet been replaced by AI Agents? Will some people snoop on their partner's interaction logs with AI Agents or LLMs to learn their deepest secrets that the partner would never share with a human? Will solving the hallucinations problem results in some users rage-switching to a more sycophantic LLM, because their current LLM repeatedly tells the user that the user is wrong and proceeds to prove why this is true?

 

[xanalog]

We had the PC and the cell/smart phone 'revolutions'.
It took us few decades and a few generations (GenX/Z) before realizing the negative-effects of tech to our young'uns.
Genβ (2025-2035) will be our truly-first digital offsprings; yet, we cannot fathom the AI's potential for damage... even from its known-unknowns!

 

[Peterinvan]

My concern is that that the powerful men that control the "ethics" while training their AI models will force the AI to produce "alternate facts" to support their push for control, power (and or profits). We could see different AI models supporting different versions of reality.

I am also totally against the anthropomorphism of the "agents". I do not accept that it makes it easier for new users to use these tools. Let's drop the "chat" paradigm.

 

[antcollinet]

Let's drop the "chat" paradigm.

Good luck with that.

 

[EERecordist]

At least one company wants to hire people to train AI to replace their jobs.

Handshake AI Fellowship | Paid, Remote AI Jobs for PhDs, Masters, and Postdocs

Earn up to $125/hr through flexible, remote projects advancing AI models with your expertise.

joinhandshake.com

 

[xanalog]

At least one company wants to hire people to train AI to replace their jobs.

Me thinx @antcollinet would readily qualify for this opportunity!

 

[sonitus mirus]

I asked for a careful, skeptical analysis, but damn.

John Bonham was almost certainly a very strong drummer, but the famous Good Times Bad Times bass triplet figure may be one of those cases where the studio version became the myth, while the live record never really seals the case. The biggest reason for doubt is simple: there is no widely known, clear concert footage showing Bonham cleanly reproducing that exact passage with the same smoothness people hear on the record. Even the best official early live documents show Good Times Bad Times mostly as an intro or fragment rather than a full, forensic demonstration. The official Zeppelin site lists “Good Times Bad Times (intro) ~ Communication Breakdown” for Paris on October 10, 1969, and the famous Royal Albert Hall film from January 9, 1970 does not include Good Times Bad Times as a standalone showcase at all.


That absence does not prove fraud, but it does support the idea that the legend rests far more on the studio recording than on live proof. If Bonham had regularly and effortlessly nailed that figure in concert, you would expect the historical record to be fuller and clearer by now. Instead, people keep pointing to partial medleys, audience tapes, and “you can kind of hear it here” moments. In other words, the evidence for the claim is oddly thin for such a famous piece of drumming.


There is also a source reliability problem. The most repeated defense of the one pedal story comes from Jimmy Page, who told Rolling Stone that Bonham could do the roll with “one foot and one pedal.” That is interesting, but Page is not a neutral witness. Led Zeppelin’s history includes multiple later credit changes and disputes over songs such as “Dazed and Confused,” “Whole Lotta Love,” “The Lemon Song,” “Bring It on Home,” and “Babe I’m Gonna Leave You.” Reuters also reported in 2025 that Jake Holmes again settled litigation with Page over “Dazed and Confused.” So while Page’s account matters, it is fair to treat it as interested testimony, not clean verification.


A skeptic can also point out that studio polish was absolutely part of Zeppelin’s method. Page has openly discussed production techniques on the debut, including double tracking on the riff in “Good Times Bad Times.” That does not mean the drum part was faked, but it reminds us that this was a crafted record, not some untouched truth serum. Once you accept that the guitar parts were being carefully built for impact, it is not irrational to wonder whether the drum legend also benefited from ideal takes, mix choices, and the general tendency of records to present the best possible version of a performance.


So the restrained conclusion is this: there is no solid evidence that Bonham did not play the part, but there is also less hard proof than the legend suggests that he could reproduce it smoothly and exactly as heard on the record. The strongest skeptical case is not “Bonham was fake.” It is: “Bonham was good, but the famous ‘Good Times Bad Times’ triplet story may owe more to studio capture, mythmaking, and weak live documentation than fans like to admit.” That position is fair, evidence based, and a lot harder to swat away.

 

[antcollinet]

Me thinx @antcollinet would readily qualify for this opportunity!

They would have to pay me a hell of a lot more than they are offering.

 

[Audio Science Pal]

I asked for a careful, skeptical analysis, but damn.

I like this part; "the general tendency of records to present the best possible version of a performance", but I did learn some interesting trivia from reading the full reply.

 

[Axo1989]

My concern is that that the powerful men that control the "ethics" while training their AI models will force the AI to produce "alternate facts" to support their push for control, power (and or profits). We could see different AI models supporting different versions of reality.

I am also totally against the anthropomorphism of the "agents". I do not accept that it makes it easier for new users to use these tools. Let's drop the "chat" paradigm.

We certainly saw this recently from xAI's Grok, which output some ridiculous Musk hagiographies responding to related and unrelated prompts.

Natural language prompts are a good part of the novelty of the generative chatbots. And the reason otherwise intelligible people appear to go nuts when writing about them. Not to mention using them.

Meanwhile: to invest in 'AI' or not? Personally I think it's an extraordinary squandering of resources and a criminal opportunity cost versus things we really need to do for human and ecosystem health and survival. On the less macro level we have some analysis from Horace Dediu:

The most brilliant move in corporate history?

In a bold departure from its Big Tech peers, Apple sidestepped the $650 billion AI infrastructure spending frenzy, a move likened to opting out of purchasing the US Navy annually. While Amazon, Goo…

asymco.com

The hyperscalers are now spending 94% of their operating cash flows on AI infrastructure. Amazon is projected to go negative free cash flow this year with as much as $28 billion in the red. Alphabet’s free cash flow is expected to collapse 90% from $73 billion to $8 billion. These companies used to be the greatest cash machines ever built. Now they’re borrowing money to keep the data center lights on.

And a corresponding story:

Exclusive: Meta planning sweeping layoffs as AI costs mount

The cuts could affect 20% or more of the company.

www.reuters.com

There's considerable unfolding to come.

 

[sam_adams]

There's considerable unfolding to come.

Remember the nonsense from Jack Dorsey at Block that GPT was saving them so much money that they could fire staff? Turns out that was corporate-speak for "We spent too much hiring people that we don't really need now so we'll blame the cuts on GPT spending." We now have a new term, 'AI-washing'. Swell.

When Block Inc. cut 4,000 jobs this week, nearly half its workforce, co-founder Jack Dorsey pointed to AI as the culprit. . . . But analysts see a different explanation: poor management.

Jack Dorsey's Block Cuts 4,000 Jobs: Critics Claim AI-Washing

The Block co-founder cut nearly half the company's staff, claiming AI made them redundant, but analysts say poor management was the real reason.

www.entrepreneur.com

https://www.bloomberg.com/news/articles/2026-03-01/jack-dorsey-s-4-000-job-cuts-at-block-arouse-suspicions-of-ai-washing

https://www.inc.com/leila-sheridan/jack-dorsey-blamed-ai-for-4000-layoffs-a-former-block-exec-says-thats-not-the-real-story/91312392

 

[sam_adams]

Oh, by the way, don't get used to 'free GPT'.

Don't get used to cheap AI

AI may never be as cheap to use as it is today. (Axios.com)

 

[tineared]

We certainly saw this recently from xAI's Grok, which output some ridiculous Musk hagiographies responding to related and unrelated prompts.

Natural language prompts are a good part of the novelty of the generative chatbots. And the reason otherwise intelligible people appear to go nuts when writing about them. Not to mention using them.

Meanwhile: to invest in 'AI' or not? Personally I think it's an extraordinary squandering of resources and a criminal opportunity cost versus things we really need to do for human and ecosystem health and survival. On the less macro level we have some analysis from Horace Dediu:




And a corresponding story:



There's considerable unfolding to come.

Perhaps AI will be the actual final frontier

Should we ask Grok?

 

[Karmacoma]

Oh, by the way, don't get used to 'free GPT'.

Don't get used to cheap AI

AI may never be as cheap to use as it is today. (Axios.com)

Remember when 25 years ago they were saying in 25 years we won't be able to afford a plane ticket because of gaz prices. Fortunately scientific discoveries almost always disrupt these kind of predictions. The cleanest energy exists, it is nuclear, fission today fusion tomorrow, it doesn't have problems of availability and has the lowest structural footprint.

 

[antcollinet]

The cleanest energy exists, it is nuclear, fission today

Clean? Ask the former occupants of Chernobyl, three mile island, Fukushima, and the sheep farmers of the lake district (admittedly a few more decades ago now)

Any process that creates some of the most toxic substances in existence as waste, that is going to remain toxic for Millenia, and that we actually have no idea what to do with for that period of time - cannot in any way be described as clean. It might be the least worst option right now (excepting renewables) but is not clean.

 

[BlueTunes]

Clean? Ask the former occupants of Chernobyl, three mile island, Fukushima, and the sheep farmers of the lake district (admittedly a few more decades ago now)

Any process that creates some of the most toxic substances in existence as waste, that is going to remain toxic for Millenia, and that we actually have no idea what to do with for that period of time - cannot in any way be described as clean. It might be the least worst option right now (excepting renewables) but is not clean.

It's an odd quirk that whilst Chernobyl did indeed dirty it's environment, in doing so and making it inhabitable for humans, it also made it super duper clean by removing that environments worst offender. (Man).

All these " energy source X is dirtier/cleaner than energy source Y" ignore the complete energy lifecycle of that particular source.

It's been postulated that both for nuclear and large scale renewables, that the total energy output never exceeds the energy input (from construction, maintenance and decommissioning).

It's a moot discussion anyway.
We should be managing our energy resources and usage to maximise the potential for unlocking true energy generation which will be free from any connection to solar input either past or present.

We are in a special place in our development. We cannot let it wither to nothing for nothing.

 

[Somafunk]

Clean? Ask the former occupants of Chernobyl, three mile island, Fukushima, and the sheep farmers of the lake district (admittedly a few more decades ago now)

Not just the Lake District, where I live/grew up in sw-scotland was badly affected by the fallout, I remember us all talking about it at school as we all watched the news and we had a local whistleblowing campaigner in the town who used to be a senior nuclear research scientist/previously worked at the Nuclear enrichment plant (for weapons grade plutonium) at the Chaplecross plant at Annan

My mates (his parents) hill farm had all its sheep culled and the area was monitored for particles for years/decades before quite recently they were allowed to graze animals again.

And our bay in the town has 11 radioactive hotspots in the mudflats thanks to all the shit what was flushed out from sellafield nuclear plant (previously windscale) before they took notice, not to mention we have the army firing range 4 miles away where 10,000+ Depleted Uranium shells were fired into the solway, with none of them ever collected as the army said it was too much work/not cost effective to remove as the saltwater will have corroded them rendering them very hazardous to move.

I'm not against it, but I want the discussion and solution regarding the waste taken into consideration before construction begins, along with how much each MW produced actually costs compared to other forms of energy

 

[Karmacoma]

Clean? Ask the former occupants of Chernobyl, three mile island, Fukushima, and the sheep farmers of the lake district (admittedly a few more decades ago now)

Any process that creates some of the most toxic substances in existence as waste, that is going to remain toxic for Millenia, and that we actually have no idea what to do with for that period of time - cannot in any way be described as clean. It might be the least worst option right now (excepting renewables) but is not clean.

"Cleanest" doesn't mean clean, it means, as you said yourself, the least worst option.

Renewables require more ressources to build, a hell of a lot more space, therefore a bigger impact on nature, have a shorter lifespan, don't have a proper energy storage solution to compensate for their constitutive output inconstancy, which if it existed would anyway produce an added gigantic amount of waste and therefore increase the already huge impact on wildlife and nature, not even talking of the estate property market.
But yes, Tchernobyl. Nuclear reactors can't explode anymore, Tchernobyl type reactors don't exist anymore, and except if you have the bright idea to build it on top of one of the most active tectonic riff hotspot on the planet, and border it with a high tsunami-risk littoral, will just melt on itself. Proper storage solutions exist for the nuclear waste, which will remain radioactive for millenials, but dangerous only for centuries, and that lifespan could, hypothetically I admit, be greatly reduced when fusion will be mastered.
Now, the countries that have really made the renewables energy choice depend on coal, basically the energy that has the biggest impact on nature and cause by far the most human deaths every year, more than dozens of Tchernobyl, because in the winter, at 8 o'clock during the peak energy demand, and there's no sun and no wind, you're just left wondering what the ''greens'' had in mind with this idea. Just not practical.
Fusion will solve all of this. That's where the money needs to go.

 

[Somafunk]

Renewables require more ressources to build, a hell of a lot more space, therefore a bigger impact on nature, have a shorter lifespan, don't have a proper energy storage solution to compensate for their constitutive output inconstancy, which if it existed would anyway produce an added gigantic amount of waste and therefore increase the already huge impact on wildlife and nature, not even talking of the estate property market.

Living in scotland I'd disagree with pretty much the entirety of your statement

 

[Salt]

"Cleanest" doesn't mean clean, it means, as you said yourself, the least worst option.

Renewables require more ressources to build, a hell of a lot more space, therefore a bigger impact on nature, have a shorter lifespan, don't have a proper energy storage solution to compensate for their constitutive output inconstancy, which if it existed would anyway produce an added gigantic amount of waste and therefore increase the already huge impact on wildlife and nature, not even talking of the estate property market.
But yes, Tchernobyl. Nuclear reactors can't explode anymore, Tchernobyl type reactors don't exist anymore, and except if you have the bright idea to build it on one of the most active tectonic riff hotspot on the planet, and border it with a high tsunami-risk littoral, will just melt on itself. Proper storage solutions exist for the nuclear waste, which will remain radioactive for millenials, but dangerous only for centuries, and that lifespan could, hypothetically I admit, be greatly reduced when fusion will be mastered.
Now, the countries that have really made the renewables energy choice depend on coal, basically the energy that has the biggest impact on nature and cause by far the most human deaths every year, more than dozens of Tchernobyl, because in the winter, at 8 o'clock during the peak energy demand, and there's no sun and no wind, you're just left wondering what the ''greens'' had in mind with this idea. Just not practical.
Fusion will solve all of this. That's where the money needs to go.

So they have found a place to bury the 'detritus' yet??