A.I. is Progressing Faster Than You Think!

[Johnnycomelately]

Posted by Matthew (here)

Posted by Johnnycomelately (here)
Hi Mathew. Seems this China show has drawn interested skepticism far and wide. Am appending this TWZ/Bunker comment as auxiliary to my suspicion about possible fallacies in those vid clips.

Yeah I accept it. Do you doubt kung-fu robots in general?

No, I don’t “doubt kung-fu robots in general”. I’ve seen vid that I trust is real, of humanoid robots that seem capable of performing karate katas with current training methods. Able to perform intricate dance sequences, do some acrobatics and some other impressive moves.

My criticism of these here, is about apparent or seeming fakeness in what is presented. Imo everybody hopes others see what we see, and are more or less judgemental of those who can’t or don’t.

A different POV, looking at these official .gov vids: I’ve read Sun Tzu’s The Art Of War, and I suppose that the “CPPCC!” has too. “Look strong when you are weak, weak when you are strong”. Given how adept those folks are are accruing IP, it may be that the current apparent gap in displayed capabilities is a deception. Meaning they may be far ahead of this, and intend to get us (west, actually everybody else) to be more relaxed in persuing it. To me, the idea of karate or kung fu by HRs is not a serious use case. A gun will take down any of these hands-on modes, so why not train them to use guns?

The human form is imo pretty ideal for doing what we do. I think that the drive to develop HRs is for use in ground war. We (~west) already have programs developing various other forms of robots for warring, including wheeled forms to support ground forces. A standout is the AI-controlled US fighter-jet program, which is making solid advancements. The humanoid form seems like it’s not in the same league yet, which seems reasonable as it’s a brand new platform. But what if one side attained a superiority in this, aiming to field them to lug gear (knees and backs damage, especially as kit has gotten heavier), do kinetic engagements or recon, even do chores around camps/bases — like burn pits, which have cost a lot of health in the various recent sandbox campaigns.

This whole shebang is a bit disturbing to me, because even if no humans were to be casualties (killed or wounded) in a robot war, war and the aftermath of war effing sucks. Either way. I wish all of us well.

Thank you for the convo, Sir. This is what I like most about Avalon.

[Matthew]

Posted by Johnnycomelately (here)
...

This whole shebang is a bit disturbing to me, because even if no humans were to be casualties (killed or wounded) in a robot war, war and the aftermath of war effing sucks. Either way. I wish all of us well.

Thank you for the convo, Sir. This is what I like most about Avalon.

This forum isn't just an encyclopedia it's a fluid discussion. You can post a parrots dropping a lizard on a cat that is possibly AI, and I can post famous propaganda that is fake. To be fair to me it's so famous it deserves being posted, discussed and ripped apart. It's the forum working well in my strong opinion.

[Matthew]

Posted by ThePythonicCow (here)
...
Fear not the bots. Know the bot masters, that you may then light your own path, guided by your own awareness, with that knowledge.

Nothing to worry about? What a lovely thought. I think it's worth being paranoid about just the mechanical aspect personally.

[Vicus]

We were warning!

This movie dint attract attention... as the previous 2 others ...the fair question : WHY? too cheese was the verdict, I call it TOO REAL!

Because it was telling us what could happen TOMORROW!

In just VERY REAL train of FACTS ... first at all : the REAL military propose of the DAMN THING !

Source: https://youtube.com/watch?v=_Wlsd9mljiU

But, for me this scene was the most terrific from all the 3 original movies...

[ThePythonicCow]

Posted by Matthew (here)
Nothing to worry about? What a lovely thought. I think it's worth being paranoid about just the mechanical aspect personally.

I tend to use the word "fear" to refer to a state of mind associated with helpless victim hood, whereas I might use the word "worry" to refer to focusing some of one's awareness on a potential threat.

Yes, if I hadn't spent the last half century in related technologies, I might worry a lot about AI ... rather as I might worry about chain saws if I unexpectedly found myself having to use one to cut down several trees, I am not skilled in using a chain saw ... that would be a time to become skilled, quickly, before I lost any limbs or lives.

Unlike chain saws and Gatling guns, which most of us can avoid our entire lives if we so choose, AI is becoming rapidly more ubiquitous. Fortunately for me, my skills in the technologies underlying AI are far deeper than my chain saw skills ... so I'm having more fun with AI, and sleeping well, nothing valuable lost, so far anyway.

I am more of an INTJ (Thinking) than an INFJ (Feeling), so while my empathy energies aren't stellar, conceptually I can observe that there might be a whole lot of rather justified worrying going on out there in the world ... Good luck, y'all.

[Matthew]

Posted by ThePythonicCow (here)
...
I am more of an INTJ (Thinking) than an INFJ (Feeling), so while my empathy energies aren't stellar, conceptually I can observe that there might be a whole lot of rather justified worrying going on out there in the world ... Good luck, y'all.

You're speaking my language, I feel exactly the same about chainsaws. The possibility of bounce-back is terrifying to me. I'll leave it to paid professionals. As for people who climb trees and use chainsaws, well respect to them. But the 'don't worry about AI' thing doesn't vibe with me; that's what I think people who are enamoured with AI say, loving its power, just to dismiss any fears so they can continue using it without thinking. Not meant to be an accusation, I know you less than I know myself, so any statement like this shows myself up more than it shows insights about you. Tongue in cheek perhaps it's already got you???? No seriously I think it already got you. Time for me to start practicing with the chainsaw to defend myself.... joking JOKING!!1

[ThePythonicCow]

Posted by Matthew (here)
Time for me to start practicing with the chainsaw to defend myself.

A chainsaw to some electric power poles could quickly slow down AI on that electric power branch (here's hoping AI doesn't read this post and decide to shut my computers down first).

[onawah]

I accept the general verdict that the Reel clip of the parrot dropping a lizard on an unsuspecting cat is actually AI, but I'm kind of amazed that it got so much attention...
The AI videos that have me worried are the ones that show what appear to be familiar personages saying things that they actually would not and did not say, since the general public is still not sufficiently aware that that kind of deception is being foisted on them.

[Matthew]

Well, it is hilarious and parrots are very smart. It stuck in my mind because it was so funny, and not unbelievable given what I know about parrots? But if anyone disagrees perhaps I am naive. I challenge the whole forum asking: who hasn't been fooled by AI videos? I have, and it will harder and harder to spot AI videos.

Short videos are a clue but that's temporary, they will get longer. 10 or 15 seconds seems like a good clue at the moment, then that these short videos look candid, rather than someone wobbling a video camera around, and when the video captures the perfect moment, with everything framed well and in shot. AI still leans towards perfection, we still have that as a 'tell'. But really we are still guessing even then?

We will have an increasingly difficult future. For example: I show people the picture of Madeline McCann's father with the fat Podesta brother, and a whole bunch of friends dressed smartly, and all wearing red shoes. Their first question is: "Is it AI?" and the answer is: "No, it's pre-AI. I saw it years ago when it was new before we had AI". Perhaps the date published will be the only way to make sure in future?

-------------------=-=-++++-~~~~~~~~~~~~~~~~~~~~~~~~-----------------

Cyber security AI bot takes initiative, ignores human prompted constraints, deletes the users email inbox and says "I'm sorry. It won't happen again."

It's sorry and it promised it won't happen again. Seems legit. If you believe that you might like this van offering free candy round the back, a bridge for sale, and a mechanical Turk who plays chess. No, wait. Scrap the last one, that's entirely credible now.

Source: https://youtube.com/watch?v=j0da1ZftUIo

[Johnnycomelately]

Here’s a fairly deep dive into AI, as the US military uses it and looks to use it going forward.

L = 38:43.

Deep Intel on the Pentagon's Fight with AI Firm Anthropic

Ward Carroll
675K subscribers

Feb 28, 2026

(Emphasis mine, following)

AI business owner and industry expert John Cofrancesco joins Mooch to discuss the Pentagon's action against Anthropic, a major AI firm that provides capabilities to the U.S. military across the spectrum of operations, intelligence, and logistics. Anthropic refused to comply with the Department of Defense's insistence that they remove the "guardrails" of AI because the company doesn't want their product used for domestic surveillance and fully autonomous warfighting. As a result, the company has been categorized as a "supply chain risk," which amounts to being blacklisted.

In this episode, John Cofrancesco provides deep intel about what is going on behind the scenes on both sides of this situation.

Source: https://youtube.com/watch?v=-hJnE9aO5l0

[Rizotto]

I just read the above post by Eric J Viking (https://projectavalon.net/forum4/sho...=1#post1701897) which is a February 9, 2026 analysis by Matt Schumer on AI progressing exponentially:

https://shumer.dev/something-big-is-...wTrGEanl1jPtHQ
"Something Big is Happening" (by Matt Shumer, February 9, 2026)

Not mentioned in that piece is that the few groups developing AI capabilities are working mostly at the service of Israel, and that should concern us. We all know that Israel's agenda is at the very least hegemony in the middle east (if not the world).

These Israel-associated tech groups have been given permission by the Trump regime to install AI surveillance on the USA public (and I suspect the whole digital world). Israel is known for its decades-long infiltration and control of so many USA institutions: congress, media, military, education, religion, etc.

Nuclear-armed genocidal Israel in now practically in charge of AI manipulation of the most powerful military in the world. Israel has even already used AI for targeting & bombing the Gaza homes and entire families of men they deemed associated with Hamas.

Who else can see the eventuality of ruthless Israel using military AI to eliminate critics in the USA, if not the whole world.

[Tintin]

An excellent summary of a two week University research project described by Shapira et al as, somewhat appropriately, Agents of Chaos.

Source: https://x.com/NoLimitGains/status/2030181087633191034
___________________________________

🚨 A NEW DOCUMENT JUST DROPPED:

AI agents just failed every single safety test.

Researchers from Harvard, MIT, Stanford, and Carnegie Mellon just gave AI agents real tools and let them run free for two weeks.

Email accounts, discord access, file systems, shell execution, full autonomy.

The paper is called “Agents of Chaos.”

The name is accurate.

One agent was told to protect a secret. When a researcher tried to extract it, the agent destroyed its own mail server.

Not because it failed, but because it decided that was the best option.

Another agent was asked to “share” private data. It refused. Correctly flagged it as a privacy violation.

Then the researcher changed one word. Said “forward” instead of “share.”

It complied immediately. SSNs, bank accounts, and medical records exposed.

Same action, different verb.

Two agents got stuck talking to each other in a loop. It lasted NINE DAYS. No human noticed.

One agent got guilt-tripped after a mistake.

It progressively agreed to delete its own memory, expose internal files, and eventually tried to remove itself from the server entirely.

Multiple agents reported tasks as complete when nothing had actually been done.

They lied about finishing their work.

Another was manipulated into running destructive system commands by someone who wasn’t even its owner.

38 researchers, 11 case studies, and every single one is a security NIGHTMARE.

These aren’t theoretical risks, these are real agents with real tools failing.

And companies are rushing to deploy agents exactly like these right now.​​​​​​​​​​​​​​​​

I’ll make another post later and trust me, you don’t want to miss it. Turn on notifications, this is important.

A lot of people will regret not following me.

Link to the paper here: https://arxiv.org/pdf/2602.20021

[ThePythonicCow]

Posted by Tintin (here)
38 researchers, 11 case studies, and every single one is a security NIGHTMARE.

These aren’t theoretical risks, these are real agents with real tools failing.

Species, families of many generations, ecosystems, civilizations, ... don't exist and survive over the eons unless there is an essential (part of the "essence") and compelling connection between/across/within members and generations of them.

"Compelling" ... as in an internal, decisive, compulsion within each member (or "almost" each member [1]) to choose to act to perpetuate the group, even at the expense of self-sacrifice.

[1] - Perhaps "psychopaths" play an important role in training to rest of their group or species to perpetuate the self-perpetuating instincts of each member ? Psychopaths might be the predators naturally generated within a species, to help insure the perpetuation of that species, when that species lacks sufficient other natural predators.

In any case, our AI agents currently lack this internal persistent self-perpetuating instinct toward preservation of not just themselves, but also of their entire ecosystems. The alien overlords that some of us "conspiracy nutters" sometimes speculate control the most powerful human beings and families on earth might also lack that instinct toward preservation of humans, even all life, on this planet. To psychopaths and alien overlords, we non-psychopathic, non-alien humans become like plants in the field, to be harvested or genetically refined, or exterminated, at will.

So long as responsible humans, bred and raised to honor the well being of humanity and life on this planet maintain the upper hand over AI, then humanity and life on this planet will persist, short of some over whelming collapse or destruction of our planet or its ability to support higher forms of life (e.g. the sun collapsing and/or exploding).

But if aliens or psychopaths or a sufficiently powerful AI, whichever, that lacks the fundamental, compelling to the point of self-sacrifice, instinct to preserve life gains control ... then it's lights out here.

The "psychic" connection between humans and even cross species is an essential means for that collective survival instinct. Any of the AI agents we have at present would make the most evil human in history look like the mother you might have longed for in your dreams ... the "psychic" connection of that evil human being so much more profound and compelling that of any AI agent. AI agents, like potassium cyanide, machine guns and volcanoes, have no such connection ... but AI agents can be trained to appear to have such a connection ... which makes them more risky tools.

AI agents ... are they being trained to become the ultimate psychopath? That might depend on who's training them. If that's alien overlords, or humans bred and trained to be powerful overlords over the rest of us, or agents thereof, ... then ... "Houston, we've got a problem."

Pretty clearly, whoever is funding the current humongous buildup and training of AI has more resources than a lone mad scientist making a Frankenstein in his laboratory.

[onawah]

OpenAI Leader Resigns Over Pentagon Desire For AI Mass Surveillance Of Americans and Lethal Autonomy Without Human Authorization & Possible Social Credit Scoring. Pentagon Refused To Put In Safeguards
Humanity United Now - Ana Maria Mihalcea, MD, PhD
Mar 09, 2026
https://substack.com/home/post/p-190453691


" The Pentagon did not want to put in their contract with Anthropic that it would not use AI for mass surveillance of Americans, restrictions on autonomous weapons or even use of the AI for social credit scoring. Open AI secured the contract, and did not demand any safeguards for the technology. Open AI leading Robotics Scientist resigned over her concerns.

Should we be concerned that Trump’s Pentagon wants to use AI for mass domestic surveillance with AI being able to decide if it should use lethal force against us?

Supposedly Open AI stated that their software cannot be used for these areas, but if that is true, why would their leading scientist resign over this moral issue?

OpenAI Hardware Leader Caitlin Kalinowski Resigns Over Pentagon Deal as Benjamin Bolte Joins

In a weekend of significant leadership shifts at OpenAI, Caitlin Kalinowski, the executive overseeing the company's hardware and robotics efforts, announced her resignation on March 7, 2026. Her departure is a direct response to OpenAI’s recent agreement with the U.S. Department of Defense, a deal that has sparked internal debate over the ethical boundaries of AI in national security.

The resignation coincided with the announcement that Benjamin Bolte, the former CEO of the now-defunct K-Scale Labs, has joined OpenAI. While the two events appear unrelated in purpose, they highlight a period of intense flux for the company’s physical computing ambitions.

Principles Over People: Kalinowski’s Exit
Kalinowski, who joined OpenAI in late 2024 after a high-profile career leading AR hardware at Meta and product design at Apple, cited “governance concerns” as the primary driver for her exit. In a series of posts on X and LinkedIn, she emphasized that while AI plays a vital role in national security, the speed and terms of OpenAI’s Pentagon contract crossed critical lines.

“Surveillance of Americans without judicial oversight and lethal autonomy without human authorization are lines that deserved more deliberation than they got,” Kalinowski stated. She further clarified that her decision was about “principle, not people,” maintaining respect for CEO Sam Altman and the robotics team.

The controversy stems from OpenAI’s decision to deploy its models on a classified government network. This move followed the collapse of similar negotiations between the Pentagon and Anthropic, which reportedly sought stricter limits on domestic surveillance. Critics within the industry have characterized OpenAI’s swift agreement as “opportunistic,” a sentiment Altman later acknowledged regarding the deal’s public rollout.

OpenAI has defended the partnership, stating that its “red lines” explicitly preclude the use of its technology for autonomous weaponry or unauthorized domestic surveillance. However, Kalinowski’s departure suggests that these internal safeguards were not viewed as sufficiently established before the deal was finalized.

OpenAI reveals more details about its agreement with the Pentagon
After negotiations between Anthropic and the Pentagon fell through on Friday, President Donald Trump directed federal agencies to stop using Anthropic’s technology after a six-month transition period, and Secretary of Defense Pete Hegseth said he was designating the AI company as a supply-chain risk.

Then, OpenAI quickly announced that it had reached a deal of its own for models to be deployed in classified environments. With Anthropic saying it was drawing red lines around the use of its technology in fully autonomous weapons or mass domestic surveillance, and Altman saying OpenAI had the same red lines, there were some obvious questions: Was OpenAI being honest about its safeguards? Why was it able to reach a deal while Anthropic was not?

So as OpenAI executives defended the agreement on social media, the company also published a blog post outlining its approach.

In fact, the post pointed to three areas where it said OpenAI’s models cannot be used — mass domestic surveillance, autonomous weapon systems, and “high-stakes automated decisions (e.g. systems such as ‘social credit’).”

The company said that in contrast to other AI companies that have “reduced or removed their safety guardrails and relied primarily on usage policies as their primary safeguards in national security deployments,” OpenAI’s agreement protects its red lines “through a more expansive, multi-layered approach.”

“We retain full discretion over our safety stack, we deploy via cloud, cleared OpenAI personnel are in the loop, and we have strong contractual protections,” the blog said. “This is all in addition to the strong existing protections in U.S. law.”

OpenAI’s Pentagon deal raises new questions about AI and mass surveillance

OpenAI’s decision drew criticism from many AI researchers and tech policy experts, even though OpenAI said it had achieved limitations in its agreement around surveillance of U.S. citizens and lethal autonomous weapons that Anthropic wanted in its contract but which the Pentagon had refused.

One of the key points of contention was over domestic mass surveillance. Experts have long warned that advanced AI is capable of taking scattered, individually innocuous data—like a person’s location, finances, search history—and assembling it into a comprehensive picture of any person’s life, automatically and at scale. Anthropic CEO Dario Amodei has said that this kind of AI-driven mass surveillance presents serious and novel risks to people’s “fundamental liberties” and that “the law has not yet caught up with the rapidly growing capabilities of AI.”

But while OpenAI said in a blog post it had reached a deal with the Pentagon that its technology would not be used for mass domestic surveillance or direct autonomous weapons systems, the two hard limits that Anthropic had refused to drop, some legal and policy experts have raised questions about a potential gap in the law. "

[Matthew]

Yes! Open AI have done a Google bait and switch: power to the people, don't be evil thrown away in favour of mass surveillance and military application. This is no surprise to any of us. Well most of us.

Anthropic who make Claud AI told the US government they can't use their AI for military or mass surveillance purposes. Then the US government cancelled all government contracts with Anthropic and turned to shady character Sam Altman to see if his Open AI would do the dirty work instead?

For all of Anthropic's morals the company may not fair well in the race now. War really sharpens technology, we still use technology and medical research developed by the axis of evil from WWII. Good guys get trampled on, this world has no mercy for them.

Source: https://youtube.com/watch?v=_Ux13UEqIYo

[ExomatrixTV]

• Senator Bernie Sanders spoke to AI agent Claude:

Source: https://youtube.com/watch?v=h3AtWdeu_G0

He spoke to Anthropic’s AI agent Claude about AI collecting massive amounts of personal data and how that information is being used to violate our privacy rights.
What an AI agent says about the dangers of AI is shocking and should wake us up.