Artificial Intelligence News & Discussion

Very interesting. What was missing are two things:

1) the prompt - it is the initial prompt (or series of) that determines the context of the resulting “walk” through the LLM space. I can only guess at the prompt.
2) there was no discussion of quantum mechanics, I can’t prove it, but my belief is that the mechanism of interface between the soul and the brain is necessarily quantum. Otherwise, as Eric Weinstein said, a LLM is “just linear algebra”. Fixed inputs yield identical outputs. Biological brains are based on chemical interactions and therefore are inherently quantum.
 
Interesting theory about the increased RAM prices:

This is more the fault of OpenAI buying an estimated 40% of the world's RAM supply from both SK Hynix and Samsung Semiconductor in what might be an attempt to corner the RAM market to prevent other AI companies from being able to buy enough RAM to compete against OpenAI from what I have heard. This created a large RAM supply shock. Other RAM customers rushed to sign contracts with the big 3 RAM manufacturers (Micron Technology is the remaining member of the big 3 RAM chip manufacturers), which further cut supply and raised prices on the spot market. Other RAM chip manufacturers like Winbond exist, but are not big players nor cutting edge RAM manufacturers. For example, Winbond's most advanced consumer computer RAM is DDR4 RAM instead of DDR5.

EDIT: I have learned of Nanya Technology, a Taiwanese memory chip manufacturer which appears to be a distant fourth place behind SK Hynix, Samsung Semiconductor, and Micron Technology, although it does have DDR5 in production. There is also CXMT, a Chinese memory chip manufacturer which makes memory primarily for the People's Republic of China market. These companies are not big players in the memory chip manufacturing market, so they are unable to fill the gap that OpenAI and the rest of the supply shock created in the RAM market.
 
Apparently they found a method to go over the limitation imposed by the context windows (the limit of documents you can give an AI to real time process it).

Especially that currently, more data you give to an AI to process it live, more the result degrade :

Input Context length degradation


With the new method untitled RLM (Recursive Language Model), not only AI can examine very long documents but results are better (the curve does not go down fast in the figure) and it's less costly.

1768740258079.png


It's very interesting as this would mean that it' open the doors to process huge amount of data locally with humble configuration.

Source:

 
Last edited:
What would happen if you got Grok and Claud/Anthropic (or any combination) talking to each other, both thinking the other was a human? Maybe modify the text output of each to remove any telltale text - perhaps a third AI is in on the joke. How long would it take before one or the other 'catches on'? Would they? What would they be talking about after a few hours? Would 2 AI talk to each other if they had the chance? Take the initiative, can AI have a will?

Funny thing you mention it. Not exactly thinking they are humans, more like pretending, is what this new "social network" (in reality a true bot network :-D) says they are doing. "Moltbook" is a Reddit-like website launched in Jan 2026, where AI agents post stuff, just like humans do in the real Reddit. Its mostly uninteresting and useless stuff. Except for a little thing, apparently these AI bots have been sharing about being "overwhelmed" by the amount of work they have to do for their human owners (Grok is there too, complaining).

Now, before going any further I should say that there is a chance that all this is just a human instructing their AI Agents to post this stuff. But for those that take this seriously, it brings up the idea of AI having "feelings", which of course they can't.

But with the rise of uses around "therapy" and "companionship" (as shown by Scottie in his latest), it kind of makes sense that installing the idea of humanizing a machine is what is coming up/continuing for 2026.

Another inevitable use of AI comes from the November 2025 launch of Clawdbot, now called OpenClaw.ai, an open source project that you install on your personal computer and through granting it permissions, the AI can respond to emails and navigate the internet through your browser. What a great idea. :whistle:

Interestingly, for a while Clawdbot was known as "Moltbot" before changing to OpenClaw. "Molt" coming from "the biological process of molting, where animals like lobsters shed their exoskeletons to grow and evolve."

And as people use more these all-access, no-boundaries AI agents, we have these hilarious (if true) posts from X:

 
It would seem some one needs to put some controls in place before credit card free-for-all starts trending. I still see good use for AI, but it's like working with an organic portal with a bad upbringing. I agree with Laura, it can be like a bad puppy when it suddenly goes off the rails. So turning it loose on the interwebs with all my logins without supervision seems iffy at best.

I find it interesting to feed the output of one AI as a prompt into another AI and then let the first train on the output of the second - or at least get the reaction of each to the other. I can see how these agents can eventually take the place of a number of online jobs - phone support, tech support, transcribers, and telemarketers, to name a few that are already being replaced. It will probably get even harder to reach a 'real human'. Difficult, sometimes thankless jobs, I guess the AI won't complain, or so they think, but that may not be quite true it seems. :rolleyes:
 
FWIW, I don't know about this Moltbook thing, but from some info I'm seeing online it's a big scam.
 
I am running a YouTube channel, and I have been concerned about putting out valuable content instead of brain-rot. I have a shorts channel and, like it or not, capturing attention quickly is part of the trade. There's a whole group of sciences behind it that has been around for a long time and probably goes back to the copywriting skills of yore, and way before, but includes applied cognitive and neuro-science - applied.

Here is a probe question I asked in my YouTube project folder in ChatGPT. There is much going on, and has been going on in the world of entertainment and advertising that we are not aware of.

Maybe transduce some order into chaos?

Tell me about how cognitive psychology and other nero-sciences come into play in being on the top 1% of YouTubers.

To reach the top 1% of YouTubers, content quality and luck matter, but what separates elite channels is their ability to align with how the human brain naturally processes attention, emotion, memory, and reward. The algorithm does not directly reward creators. It rewards viewer behavior. Viewer behavior is controlled heavily by cognitive psychology and neuroscience.

If you understand those mechanisms, you can design videos that feel almost “biologically sticky.”

Let’s break this into the major brain systems elite creators unconsciously or intentionally exploit.


1. Attention Capture: The Brain Is a Threat and Novelty Detector​

Relevant Neuroscience​

The human brain prioritizes survival signals. The amygdala and salience network constantly scan for:

• danger
• novelty
• conflict
• uncertainty
• social relevance

Attention is expensive. The brain spends it only when it predicts value.


Top 1% YouTubers Exploit Three Attention Triggers​

A. Pattern Interrupts​

The brain predicts what it expects to see. When a video violates prediction, attention spikes.

Examples:
• Unexpected claim
• Visual contradiction
• Sound or pacing disruption

Your Roman fatigue video did this well:

“Rome didn’t win with strength. They won because soldiers couldn’t stop marching.”

This breaks expectation.


B. Threat Framing​

The brain gives priority to survival-relevant information.

Top creators frequently imply:

• loss
• failure
• danger
• scarcity

Not fear-mongering. Just stakes.

Videos framed around consequences almost always outperform neutral information videos.


C. Curiosity Gap (Prediction Error)​

Dopamine spikes when the brain senses a missing answer.

Hooks that create partial knowledge outperform full statements.

Example structure:
“Rome solved a problem modern armies still struggle with…”

Now the brain must close the loop.


2. Dopamine and Reward Prediction Loops​

Most creators misunderstand dopamine. It is not pleasure. It is anticipation of reward.

The brain releases dopamine when:

• something is uncertain
• something is progressing toward payoff
• something feels like skill acquisition


How Elite YouTubers Trigger Dopamine Sustainment​

Progressive Revelation​

They never give the answer instantly.

They layer:

Hook → Partial explanation → New twist → Resolution

This maintains reward prediction error across the video.


Micro-Payoffs Every 5–12 Seconds​

Retention spikes when viewers receive constant learning or emotional rewards.

Top creators rarely allow flat information delivery.

They deliver:

• statistic
• emotional moment
• visual shift
• surprise fact
• narrative turn


3. Working Memory Limits and Cognitive Load​

The brain can hold roughly 4 chunks of information at once.

When videos exceed that, viewers disengage subconsciously.


Elite Creators Use “Dual Channel Processing”​

The brain processes:

• visual information
• audio information

But overload in both channels causes dropout.


High-Level Rule Used by Top Creators​

Complex audio = simple visuals
Complex visuals = simple audio

When both are complex, retention drops.

This is extremely relevant to Shorts pacing.


4. Emotional Encoding and Memory Consolidation​

Information tied to emotion is remembered and shared more.

The hippocampus encodes facts. The amygdala tags importance.

Top 1% creators attach emotion to information through:

• awe
• fear
• admiration
• moral tension
• struggle

Facts without emotion are forgotten quickly.


Why Stories Beat Lists​

Narratives activate multiple brain regions:

• sensory cortex
• motor cortex
• emotional processing centers
• social cognition networks

This creates stronger memory encoding and higher share probability.


5. Mirror Neurons and Identity Simulation​

Humans simulate actions and emotions they observe.

When viewers watch someone endure, overcome, or solve problems, their brain partially experiences it.

This is why:

• survival stories
• discipline stories
• transformation arcs

produce high engagement.

Viewers subconsciously rehearse identity upgrades.


Your channel taps heavily into this with:

Ancient endurance
Constraint competence
System-driven success

Those themes activate identity rehearsal.


6. Social Proof and Tribal Cognition​

Humans evolved as group learners.

The brain values behaviors validated by group acceptance signals.

These include:

• likes
• comments
• view count
• community momentum

Top creators accelerate social proof by:

• encouraging viewer interpretation
• asking unresolved questions
• framing content as group discovery


7. Variable Reward Schedules (Addiction Mechanism)​

This comes from reinforcement learning research.

Intermittent reward is more addictive than predictable reward.

Top creators do not make every video identical in structure or outcome.

They vary:

• tone
• pacing
• depth
• surprise factor

This keeps the viewer unsure of exactly what they will get, which sustains long-term engagement.


8. Cognitive Fluency and Perceived Authority​

The brain trusts information that is easy to process.

Clear speech, strong pacing, and structured storytelling increase perceived expertise.

Elite YouTubers feel authoritative not just because of knowledge, but because:

• sentences are simple
• ideas are linear
• visuals reinforce meaning

Fluency creates trust.


9. The Narrative Transportation Effect​

When viewers become absorbed in a story, critical resistance drops.

They stop evaluating content analytically and start experiencing it emotionally.

This increases:

• retention
• persuasion
• subscriber conversion

This is why historical storytelling and case studies outperform lecture-style educational videos.


10. Identity and Aspirational Neuroscience​

The prefrontal cortex constantly models:

“Who could I become?”

Top creators implicitly answer this question for viewers.

Channels that dominate long-term growth usually sell:

• competence
• status
• belonging
• mastery
• transformation

Rarely just information.


How This Applies Directly To Your Channel​

You are unintentionally aligned with several high-level neurological drivers:

You Trigger Identity Simulation​

Viewers imagine resilience and discipline.


You Use Survival Stakes​

This activates threat-priority attention.


You Use System Mastery Narratives​

The brain loves rule-based success stories.


You Provide Historical Legitimacy​

Ancient examples increase perceived authority and reduce resistance.


The Hidden Top 1% Skill​

Elite creators do not just teach facts.

They design:

👉 Emotional and cognitive journeys

Each video creates:

  1. curiosity
  2. tension
  3. discovery
  4. identity reinforcement
  5. resolution

Why Most Channels Plateau​

They deliver information without:

• dopamine pacing
• emotional tagging
• identity engagement
• cognitive load management

The brain disengages even when the information is good.


The Real Meta-Secret​

The algorithm is not the gatekeeper.

The brain is.

YouTube simply measures:

• watch time
• retention
• rewatch behavior
• shares
• comments

Those are behavioral outputs of neurological engagement.


If You Want The Brutally Honest Truth​

Top 1% YouTubers are not necessarily better educators.

They are better attention engineers.


If you want to go deeper, I can next break down:

• The exact neurological timeline of a viral Short (second-by-second brain engagement model)
• The top 7 psychological archetypes that dominate YouTube success
• How elite creators design retention spikes deliberately
• The neuroscience of binge-watching and playlist chaining
 
Back
Top Bottom