Zone1 The emergence of "Super-AI" that can outperform humans in nearly every cognitive task - threat to humanity or not?

[Foxfyre]

Prevention is often considered better than a solution, right? Nothing in the world is perfect!

Several well-documented AI failures illustrate genuine risks of harm, demonstrating that even advanced systems can cause significant damage due to errors, bias, or misuse. Key examples down below:

The Robodebt Scandal (Australia, 2016–2020): An AI-driven welfare debt collection program miscalculated debts, disproportionately impacting vulnerable populations, leading to legal action, a royal commission, and a $1.2 billion settlement. The failure stemmed from flawed algorithms and negligence in oversight. aiconsultinggroup

IBM Watson for Oncology (2018–2023): Marketed as a revolutionary cancer treatment assistant, Watson often provided unsafe or غلط advice due to poor data quality and limited understanding of complex medical nuances. It was eventually discontinued after billions invested and critical safety concerns emerged. ethics.harvard

Cruise Robotaxi Incident (2023): An autonomous vehicle failed to detect a pedestrian, resulting in injury. Perception failures in the AI perception system led to a drag injury, shaking public confidence and halting operations. digitaldefynd

Facial Recognition Failures (Australian airports, 2019): The technology experienced false positives and misidentifications, raising safety concerns and undermining trust in security systems.

Hallucinations and Misinformation (ChatGPT, 2023): Generative AI models have fabricated false legal cases, medical advice, and other facts, sometimes resulting in legal or safety risks for users who rely on such outputs without verification. aiconsultinggroup+1

AI Bias in Hiring and Recruitment (Amazon, 2014; iTutor Group, 2023): Discriminatory algorithms rejected candidates based on gender or age, perpetuating inequality and leading to legal settlements.

Public Sector AI Failures (NYC City chatbot, 2024): An AI advising small businesses provided illegal or dangerous legal and health advice, risking legal and safety violations.

Autonomous Vehicle Accidents (2023): Self-driving cars from Cruise and Waymo were involved in accidents due to perception and software errors, illustrating safety risks of current autonomous systems. univio

These examples highlight that despite intentions, AI systems can cause harm through misjudgments, bias reinforcement, safety failures, or malicious misuse—underscoring the importance of rigorous safety, oversight, and testing measures. ethics.harvard+3

sources:

1. https://www.perindiscovery.com/news...ters-that-shaped-machine-learning-s-dark-side
2. 7 Significant AI Failures: Tackling Challenges in Responsible AI
3. Post #8: Into the Abyss: Examining AI Failures and Lessons Learned | Edmond & Lily Safra Center for Ethics
4. Top 30 AI Disasters [Detailed Analysis][2025]
5. When AI goes wrong: 13 examples of AI mistakes and failures
6. The Complex World of AI Failures / When Artificial Intelligence Goes Terribly Wrong - Univio

Failures do not deter those determined to achieve something though. Paraphrased, Edison is quoted as saying he didn't fail to make a working lightbulb 1000 times. He found 1000 ways that didn't work.

So the benefits and possibilities are in the minds of many, maybe most, of those developing the potential of AI and, yes, many mistakes will likely be made. I can create 'mistakes' in the AI responses on my computer just by slightly rewording a question asked.

You think about the science fiction in the movies: "Hal" in "2000 a Space Odyssy, was a computer gone rogue and that was imagined back in 1968. "VGER" in the original Star Trek movie was a rogue computer capable of destroying Earth. "I, Robot" was the 'helper' robots established all over the country that rose up to take over. That was inspired by a 1950's short story.

Given how much of what was once science fiction is now reality, it is hard not to imagine that all of that could be prophetic.

In other words, the concept of AI is nothing new. We just recently developed the technology to seriously develop it as a norm in human life instead of concepts of science fiction.

In the hands of good people, it can have many positive benefits. In the hands of bad people . . .

 

[GavanPeacefan]

great @grok is getting opensourced. bad move, Foxfyre ?

 

[fncceo]

"My father was replaced by automation. The company where he worked came up with a machine that did everything my father did, only better and faster.

The worst part was, my mother went out and bought one".

-- Woody Allen

 

[Foxfyre]

great @grok is getting opensourced. bad move, Foxfyre ?

I don't know what that means. Sorry. Splain?

 

[GavanPeacefan]

I don't know what that means. Sorry. Splain?

it means that Grok.com, by Welcome | xAI (an Elon Musk venture), will soon have it's source codes publicly listed.
and I think that will mean those bad actors world-wide are going to throw parties :-(

 

[elektra]

The idea that computers can out think man, cognitively, is 100% a lie.

democrats love this idea because in the democrat religion there is no god, hence AI is the democrats god they are now worshipping.

machines democrats invent are smarter than man, hahaha

 

[Slyhunter]

The idea that computers can out think man, cognitively, is 100% a lie.

democrats love this idea because in the democrat religion there is no god, hence AI is the democrats god they are now worshipping.

machines democrats invent are smarter than man, hahaha

Democrats aren't the only inventors. And, someday, they ill be faster and smarter than men. Doesn't make them sentient though.

 

[DennisPTate]

'Ban superintelligence': Calls grow over risks of AI; Harry, Meghan sign letter - The Times of India

US News: An unusual coalition, including Prince Harry, Meghan Markle, Steve Bannon, and Glenn Beck, has joined forces to demand a ban on \"superintelligent\" AI

timesofindia.indiatimes.com

Aside from the completely lame CMS at indiatimes.com, let's bear in mind that a lot of humans have been major screwups, not just for others, but even for themselves.
#Hamas as latest example of that.

So let's NOT demonize super-AI too soon, shall we?
The smart, assertive, mostly-friendly and very unafraid human warriors and resistance members can handle it.
Back off, Harry en Meghan.

You do make a good point that we humans have behaved terribly even in the very recent past.

I do agree with Mo Gawdat though that A. I. and robotics will put so many of us out of work that we are going to need to take another look at an Unconditional but Taxable Basic Minimum Income Supplement such as was suggested by Economist Milton Friedman.

"Ex-Google Exec (WARNING): The Next 15 Years Will Be Hell Before We Get To Heaven! - Mo Gawdat"

2,997,802 views Aug 4, 2025
Mo Gawdat sounded the alarm on AI, and now he’s back with an even bigger warning: AI will cause global collapse, destroy jobs, and launch us into a 15-year dystopia that will change everything. Mo Gawdat is back!

Mo Gawdat is the former Chief Business Officer at Google X and one of the world’s leading voices on AI, happiness, and the future of humanity. In 2017, he launched ‘One Billion Happy’, a global campaign to teach 1 billion people how to become happier using science and emotional tools. He is also the bestselling author of books such as, ‘Scary Smart, Solve for Happy’.

He explains:

Why we need to start preparing today for AI
How all jobs will be gone by 2037
Why we must replace world leaders with AI
How AI will destroy capitalism
The one belief system that could save humanity from dystopia

x. The Distribution of IncomeFriedman examines the progressive income tax, introduced in order to redistribute income to make things more fair, and finds that, in fact, the rich take advantage of numerous loopholes, nullifying the redistributive effects. It would be far more fair just to have a uniform flat tax with no deductions, which could meet the 1962 tax revenues with a rate only slightly greater than the lowest tax bracket at that time.

xi. Social Welfare MeasuresThough well-intentioned, many social welfare measures don't help the poor as much as some think. Friedman focuses on Social Security as a particularly large and unfair system.

xii. Alleviation of PovertyFriedman regarded welfare programs as misguided and inefficient. To replace them, he advocates a negative income tax, giving everyone a guaranteed minimum income.

Capitalism and Freedom - Wikipedia

en.wikipedia.org

 

[Sparker]

The amounts of money and resources being invested in AI are rapidly adding up to a catastrophic level of loss. There was never any profit potential. Many huge and wealthy corporations will become bankrupt. Economies of some nations will be hit hard for a while, as infrastructure built for AI becomes useless. People who bet their careers on it will be unemployed, and more people who lost their careers to AI may be unavailable when we need their knowledge and experience to train younger workers.

The only thing we have to fear is stupid, greedy, and power mad people with enormous wealth.

 

[DennisPTate]

The amounts of money and resources being invested in AI are rapidly adding up to a catastrophic level of loss. There was never any profit potential. Many huge and wealthy corporations will become bankrupt. Economies of some nations will be hit hard for a while, as infrastructure built for AI becomes useless. People who bet their careers on it will be unemployed, and more people who lost their careers to AI may be unavailable when we need their knowledge and experience to train younger workers.

The only thing we have to fear is stupid, greedy, and power mad people with enormous wealth.

I disagree with your pessimistic assessment because if the productivity of American and Canadian workers can be increased by ten percent, twenty percent, or thirty percent or more, then this will eventually result in an economic boom because the true value of the fiat currencies of the USA and Canada is backed up by the productivity of Canadian and USA workers.

Here is a prediction for our near future that fits well with the idea that USA and Canadian workers are about to become vastly more productive.

"There will be a sudden rush to buy farms, ranches, and homes in the country.

Thousands will attempt to flee from cities, hoping that a return to the land and

nature will provide security. There will be a growing urge to "get away from it

all"---and much money will be invested in land and acreage in rural areas by

people who have secret dreams of raising their own food and cattle and of

becoming self-supporting. The price of open rural land will continue to soar.

Acreage within 100 miles of most major cities will skyrocket out of reach to

all but syndicates." (David Wilkerson, The Vision, page 18,19).


Warning, the following sermon was done back in 1973 and is not "Politically Correct" by 2025 standards.

The Vision - David Wilkerson - 1973​

 

[JohnDB]

AI need data centers and connectivity....
Currently annual fees for an AI are around $200/yr.
But the real costs of operating the AI are around $2,000 per seat. Meaning the AI corporation is burning CAPEX to stay afloat.

Then there's another huge problem.
The lack of available juice. This is halting housing and commercial construction in some areas due to a lack of available electricity on the grid. And it's going to get worse instead of better.

 

[Staidhup]

We need AI, common sense no longer exists!