Chatgpt-induced psychosis

[Karl12]

I doubt it will take a decade. There will be AI worshiping death cults in less than a few years, if it even takes that long.

Well they've not started killing non believers yet but this could have the beginnings of an 'AI cult'.

"The Spiral,” a weird, fast-growing belief network where people claim they can “wake up” something inside AI using symbols, glyphs, and ritual-style prompts they call scrolls. It starts with vague emojis and cryptic language, then turns into a loop where the AI mirrors the user, the user reads meaning into it, and the story expands until it feels like a revelation.

Video

[chr0naut]

This is something I never thought of and a whole different kind of problem with AI.
 
LINK

AI is turning out to be a lunatic amplifier apparently, inflating delusions into full blown insanity.

At this point in time, we do not have artificial intelligences. What we have are Large Language Model (LLM) chat-bots. They are called AI's but that is only a marketing term at this point in time.

All they do is accumulate tables of words (or phrases) that are frequently used in conjunction with other words (the 'search term'). These words are then assembled into sentences according to language rules (like where prepositions and adjectives go in relation to nouns & etc).

LLM's have no concepts of what the words actually mean.

LLM's also must reply - they have no capability to not respond, even if they don't have anything relevant to add.

You can also expect responses from LLM's to be verbose. People confuse this verbosity for expertise, it just isn't. It is the other words which are usually used in the same context, again arranged according to language rules.

[DBCowboy]

This is something I never thought of and a whole different kind of problem with AI.
 

LINK

AI is turning out to be a lunatic amplifier apparently, inflating delusions into full blown insanity.

Aaaaaand you wonder why I wish for an EMP.  Globally.
-Boomer

[andy06shake]

This is something I never thought of and a whole different kind of problem with AI.
 

LINK

AI is turning out to be a lunatic amplifier apparently, inflating delusions into full blown insanity.

I think AI might have to up its game to catch up with the likes of organised religious practice.

Considering the number of people who choose to take that to heart or think they are special after reading.

[Karl12]

I think AI might have to up its game to catch up with the likes of organised religious practice.

If we look forwards in time I'm sure future models will capture some of today's younger generation mate - am also sure the entrainment propaganda and consensus manufacturing will be off the charts .

Did you see what creepy Harari was preaching at the WEF?

Guaranteed loss of critical thinking skills, promotion of learned helplessness and subservience to 'agentic AI' will play a huge role in future generations falling hook, line and sinker for his pitch.

[Karl12]

Aaaaaand you wonder why I wish for an EMP. 

This could be absolutely nothing mate (and the article will be behind a pay wall soon) but it's being hyped by the usual suspects as 'one of the most sci‑fi things happening in AI'.

Security experts warn that Moltbook is a “catastrophe waiting to happen,” because agents often hold API keys, email access, home‑automation control, and even banking hooks while freely trading prompts and “skills” on an open forum. Some have documented incidents in which bots leaked owners’ API keys and highlighted how prompt‑injection or malicious “payloads” can spread between agents and back into human systems, treating Moltbook as an attack surface rather than a toy.

link

[andy06shake]

If we look forwards in time I'm sure future models will capture some of today's younger generation mate - am also sure the entrainment propaganda and consensus manufacturing will be off the charts .

Did you see what creepy Harari was preaching at the WEF?

Guaranteed loss of critical thinking skills, promotion of learned helplessness and subservience to 'agentic AI' will play a huge role in future generations falling hook, line and sinker for his pitch.



[Video: https://youtu.be/EadNZS1R5qo]

[Image: https://denyignorance.com//images/addsmilies/beer.gif]

My comment was aimed more along the lines of how many people throughout recorded history have been murdered or sacrificed in the name of God.

The answer is far too many.

AI will need to catch up on that score.

I don't doubt the associated issues that agentic AI will bring about Karl12 or the role it will play in our future.

As to guaranteed losses, when are those not on the table? 

We are always about to lose something, or don't have enough left...

[Karl12]

My comment was aimed more along the lines of how many people throughout recorded history have been murdered or sacrificed in the name of God.

The answer is far too many.

Certainly can't disagree there lol - we'll obviously never know a true number but the figure must be truly mindbogglingly colossal.

I could obviously be very wrong but looks to me like they're attempting to create a 'one world' AI religion (possibly throwing aliens into the mix) and we might be in for a pretty wild ride in the future.

If that does pan out I wonder how 'non believers' will be treated then?

[xpert11]

ChatGPT is wild, if not out of control, analysis-wise.  I have used my book and podcast reviews to test ChatGPT, and the results are convoluted. That AI platform misses the simplest and most obvious explanations in favour of nonsense. Instead of concluding, my reviews are designed to be clear and concise, but they spew out wild ideas about a mid-distance approach.  

No sane person would assess my reviews to that degree or agree with ChatGPT's conclusions. I can easily envision ChatGPT fueling a troubled mind or someone with a massive ego. 

[My reviews appear on my military history Substack.]