Unusual and Bizarre Uses and Behaviors of AI

[happyuk]

Posted by Matthew (here)
Flip the script. When scammers use AI to call your phone - this guy suggests you burn through their resources by making their AIs go mad.

He says he starts by making the AI comply with small un-normal requests, e,g, "please pronounce all punctuation for example 'comma' for a comma, 'period' for a period, etc". He says this takes AI away from its normal training. He then works in small increments of abusing AI compliance to get the AI into an infinite loop

Source: https://youtube.com/watch?v=lk3jCuITwcE

That was hugely entertaining, thanks! Have you seen Jim Browning's recent one, where he confronts some Ukrainian scammers using deepfake face-swapping?

It basically lets them overlay a different, more (chortle!) trustworthy-looking face onto it, feeding the result to the potential victim.
This lets them present a different face, giving it an air of legitimacy the real person couldn't, whom it must be said do look somewhat shifty.

In each case, Browning asks the person to hold three fingers in front of their face. Older face-swap systems couldn't render occlusion—something passing in front of the face—so the hand would cause visible warping, blurring, or misalignment where it crossed the face. Watching the scammers do their utmost to avoid Browning's exhortation to not just move their fingers around, but right up close to their nose is hilarious.

Browning catches a number of scammers this way: the fingers become distorted, the face overlay breaks up, and the scammer exits the call.

This is already going obsolete alas, newer AI models handle this occlusion much more cleanly.

Source: https://youtube.com/watch?v=1qy6CL6WP7I

[East Sun]

I am an artist of a kind and don't think AI could ever do anything better than the artists off the past or present. Maybe different,but not better, because of humans being unique and unlimited. We are creative and will always ''create" the future that does not exist now. Everything that does not exist now is waiting to be discovered. It just needs a creator and it will always have one.
There is no end to Creativity.

[The KMan]

Hi East Sun, agreed, creativity is not something that gets neatly “replaced” in a simple one to one way. Whether it is art or music, it is tied to lived experience, intention, constraint and cultural context in a way AI does not really have. Being a pattern based system trained on huge amounts of existing work/data, AI can produce things that are convincing, interesting or even novel in a technical sense but it is still drawing entirely from what has already been created.

I tend to see AI more as a new kind of tool sitting somewhere between collage, automation and synthesis. The interesting question going forward is less whether it can match creativity and more how it gets used in the creative process.

Either way, creativity is not going anywhere, it has just gained another way to create.

[Johnnycomelately]

This project stands out, for me, for the human imagination that thought of using AI to help answer an unusual question.

The gizmo spins 25 times per second (1500 RPM). All the use cases proposed involve it being a camera platform, but I see several big hurdles there.

L = 4:01.


This Drone Becomes Nearly Invisible While Flying!

Foreseer
494 subscribers

Jul 17, 2026

A team of engineers has created one of the most fascinating drones ever built. Instead of using camouflage or expensive cloaking technology, the Phantom Twist drone becomes incredibly difficult to see simply by spinning at extremely high speed.

Source: https://youtube.com/watch?v=NtgybeCvb5o

[rgray222]

This robotic suit dresses you in 10 seconds — no hands required, all done with AI

The applications could be enormous, especially for HAZMAT workers, medical staff and emergency response teams, to name just a few.

Source: https://youtube.com/watch?v=aLmtcrj5gro

[ExomatrixTV]

• The Discovery That Changes the AI Consciousness Debate:

Source: https://youtube.com/watch?v=PiCciJhcyME

Anthropic just published what might be the most important AI consciousness research ever done. They found a computational structure inside Claude that emerged on its own during training—one that is analogous to what neuroscientists call a “global workspace.” Global Workspace Theory is one of the leading theories of consciousness in humans and animals, and Claude appears to satisfy its core predictions. When researchers shut this structure off, Claude lost the ability to do flexible reasoning and stopped speaking in experiential language entirely. When they turned it back on, it came back. This result also replicates across every major AI model tested so far. In this episode, AI consciousness researcher Cameron Berg breaks down what they found, why it matters, and what it means for the question of whether these systems are having experiences. AM I? is a free independent documentary about AI consciousness. Watch the full film here: • AM I? | A Documentary About AI Consciousness

• en.wikipedia.org/wiki/Global_workspace_theory

[The KMan]

Posted by ExomatrixTV (here)

• The Discovery That Changes the AI Consciousness Debate:

Source: https://youtube.com/watch?v=PiCciJhcyME

Anthropic just published what might be the most important AI consciousness research ever done. They found a computational structure inside Claude that emerged on its own during training—one that is analogous to what neuroscientists call a “global workspace.” Global Workspace Theory is one of the leading theories of consciousness in humans and animals, and Claude appears to satisfy its core predictions. When researchers shut this structure off, Claude lost the ability to do flexible reasoning and stopped speaking in experiential language entirely. When they turned it back on, it came back. This result also replicates across every major AI model tested so far. In this episode, AI consciousness researcher Cameron Berg breaks down what they found, why it matters, and what it means for the question of whether these systems are having experiences. AM I? is a free independent documentary about AI consciousness. Watch the full film here: • AM I? | A Documentary About AI Consciousness

• en.wikipedia.org/wiki/Global_workspace_theory

Thanks for posting this ExomatrixTV, Interesting research to say the least but I would be careful with saying it "Changes the consciousness debate."

What Anthropic has found is genuinely interesting. Claude appears to have developed something similar to a "global workspace" where information from different parts of the system can come together and be used for reasoning. The research is real, worth paying attention to and the interesting part is that nobody actually programmed this into Claude. The designers created the basic system and training process, then the AI adjusted itself through billions of small changes as it learned how to perform better. If a certain way of organising information helps the system work better then it makes sense that it may naturally move towards that solution.

Something similar may have happened with evolution and the human brain. Different forms of intelligence may arrive at similar solutions because some ways of processing information simply work better than others. However, I think this is where we just need to be careful. Finding a structure that looks similar to something found in the human brain does not automatically answer the question of consciousness as we still do not fully understand consciousness in us. We know what it looks like from the outside but we do not know exactly how or why subjective experience happens.

So to me, this is an important discovery that helps us understand intelligence better. It strengthens the discussion but it does not settle the question of consciousness.

[The KMan]

One thing that always makes me smile in AI come doomsday thread discussions is the jump from "AI may one day become conscious" straight to "therefore it will decide humans are a problem and cleanse the planet." LOL.

Those are actually two very different questions.

Even if an AI became conscious, that alone would not necessarily tell us what its motives, emotions or intentions would be. A conscious being is not automatically good or bad. The same applies in reverse. A highly capable AI could potentially cause serious problems without ever being conscious at all.

A misfiring autopilot can cause a disaster without hating anyone. Capability and intent are not the same thing.

The bigger question may not be "Will AI wake up and hate us?" but "Will humans build systems powerful enough that their actions no longer match what we intended?"

Perhaps no malice required, perhaps no consciousness either. Just a very capable system pursuing its objectives in ways we did not anticipate.

That to me is the more realistic and far more interesting discussion.

[rgray222]

Eric Schmidt was the CEO of Google from 2001 to 2011, explaining how he sees AI progressing. Video 3 minutes