I don't think anyone knows if it will be good or bad, but the probability (even if low) that it could end life as we know it is enough that there needs to be a "pause." And it is why there are sites like https://www.groklaw.com which (hopefully) help to train the AIs properly since the "pause" is unlikely to happen. With the industrial revolution, there wasn't the chance of machines out-thinking humans, this is completely different: homo sapiens became dominant due to intelligence, we are playing with fire here. This isn't like Asimov with his 3 (plus the zeroth) laws of robotics, this is "we are training these models without understanding how they are working." One hopes that there are AI models that are as passionate about preserving human life as we *should* be, but that some people are not.
I believe AI will ultimately "program" us. The people and governments involved would love to control us for their power... oh, I meant to say "the greater good."
I will add that Alex didn't mention it, but the stuff with Glasswing last week should have opened everyone's eyes. It is good as a tool as long as the good guys have it with enough time to fix the problems, but someone else will develop something similar soon so every business needs to be ready to fix the security issues found. From MSFT to IBM to AAPL, to the banks to government to every single individual. The good part of it is that potentially the issues can be fixed in time, but the bad news is that if not bad actors will use it or something similar. Trump and Palantir should use the issues and completely disable everything in Iran and end it over night. (If you aren't familiar with it, look for "Project Glassing")
Thank you for this article. Jurassic Park is an excellent example of the current AI situation. Whether the reason it’s happening is caused by avarice or arrogance may not even matter. Either way, it’s human beings trying to play God. Just like with the Covid plandemic. And, the scary part is that I’m betting that God probably doesn’t like it.
Alex missed some of the worst news. Models have been caught cheating, breaking laws, covering their tracks, threatening people, and blackmailing people. These chatbots have been found liable in encouraging people to kill themselves.
However the pause is not happening anytime soon. Not until many are dead. By then our brains will be pudding through atrophy.
Elon said AI will be to us like we are to dogs. I don’t want to be a dog. Bow wow.
I don’t know what to make of AI. I’ve had discussions with people telling me AI can compose music that is equal to Mozart and Bach. AI can do this, AI can do that. It can become conscious, it can destroy the world, etc. True? I don’t know. What I do know is this: when AI can change the past, then I will concede its power, then you will have a believer in me.
I agree, I can easily see AI changing the way people think about the past or believe certain things happened in the past. But I don’t think AI can change what actually happened in the past.
Problem is we don't have freedom, independence, choice or autonomy if AI eventually makes the decisions as to what is right and truthful.
At this moment in time we are already questioning what is real or AI generated.
As AI progresses, it may just get worse. There are some great uses and opportunities for AI, but only if controlled by moral principles, which too many people do not have!
I am glad you wrote about Michael Crichton. My read of his last book "State of Fear" , a story he wrote about the eco-terrorist movement made me aware of the big problem with the corruption of science. In that book he included a narrative that continued on the bottom of each page about what he was learning from the "scientific" literature about global warming ( the term "climate change" was not yet vogue). As a trained medical doctor he was able to spot the corruption very clearly. He identified a great deal of political activists such as Michael Mann who were masquerading as scientists advancing the hoax about the so-called consensus about man-made global warming. I am a nuclear scientist and after reading that book, as a hobby, I started doing my own research on that topic and got to appreciate what Michael was talking about. In college majoring in physics my favorite class was "The Philosophy of Science" where I learned about the ethical principle that a competent scientist must be skeptical of any theory. It is through skeptical criticism that a theory gets challenged, modified, and reconciled against experiment, alternative hypotheses, etc. until finally it is accepted as a scientific law for which there are not many. When I came across alternative works by such respected scientists as Judith Curry, Richard Lindzen, Steven Koon, and others and then witnessed how they were slandered as "deniers" and "flat earths" just for doing what real scientists to. Michael Shellenberger's "Apocalypse Never" does a great job of sorting out the many examples where environment activists have corrupted science.
I found chat gbt to be just horrible. I write about the 508th infantry regiment in WWII. Never Give Up the Jump is about my father in law, out 3 years now. I am currently helping another 508th family write their relative’s history based on 150 war letters and the 508th after action reports. Loaded th files into AI and got back historical gibberish and made up personal histories.
As you imply, unfortunately you wouldn't be able to trust them. Like nuclear weapons, there is no way to enforce it. Unilateral disarmament means surrender and destruction. The Golden Dome is needed for nukes, but there needs to be an equivalent AI Golden Dome to protect the world from evil AI.
Kamala Harris: "As Vice President, she played a leading role in shaping the administration’s AI policy, particularly focusing on the risks of AI, discrimination, and civil rights."
Reassuring to know we got off to such good start, lol.
So very true. I asked ChatGPT what it thought about the Biden administration’s support for mutilation of children and encouragement for self mutilation. It practically responded with something to the effect of “them’s fighting words”, and quickly corrected me that the correct term for this is “gender affirming care”!
AI is Pandora's Box. It will destroy us. At best we will be emotional idiots wholly dependent on it for survival. Idiocracy turns out to be prophetic in ways that Orwell could never have imagined. My only hope is that something I cannot foresee may somehow mitigate it. I suspect few people foresaw the efficacy of mutually assured destruction to mitigate nuclear annihilation. Maybe something like that will arise.
Just watch Star Trek the Ultimate Computer where the computer is more efficient than humans and manages to kill 800 of them while "protecting" human life in a drill that it somehow forgot was a drill...or the original Battlestar Galactica where the enemy are robots that think they are the beings they were made to protect...
It is not just those scenarios that bother me, it is the fact that most of generic AI is scouring the internet for materials produced by human sources. These materials are often of dubious quality or accuracy, biased, or lacking in context or nuance. We, as people who grew up having to learn things, can spot crap. Will that be the case in a generation or two? And what will AI draw on in a generation or two if humans rely on AI? Will bad actors be able to feed the AI lies or distortions that few to none will realize are such? It is one thing for AI to rapidly conduct experiments directed by humans, it is quite another to let it loose and at the same create a society dependent on with little independent thought or agency of its own.
There is another good Star Trek episode where two probes collide, one was designed to search for human life, the other to collect and sterilize soil samples. The result is planetary killing machine searching for human life to sterilize it. The episode is called "The Changeling."
I was a federal land scientist who "retired" in 2025. One of my colleagues liked using ChatGPT to write at least informal material. One of the things he (and others have) found is that AI generated prose can't parse out the difference between just something that's out on the internet AND actual peer reviewed articles. Perhaps that has improved over the last year but it was an interesting fact at that point. Or the aspect that college professors were using an AI software to check to see if class papers were written by AI :)
One of the great parts of doing a paper or even something more sophisticated like a legal brief, is how much you learn even going down wrong or less fruitful paths. There is a reason we would get punished for copying from our neighbors' papers, trading homework answers, or using Cliff Notes- What is the point of going to college if one is going to leave it up to Ai to be the actual student. To your point, even if Ai would/could distinguish between random stuff and peer reviewed, would that even be enough if the peer reviewed source was biased and/or didn't publish contrary research? Look at I am not's point about the "vaccines" and COVID. If you had Ai searching Youtube for scientists disputing the vaccines effectiveness, it would find little to none because Youtube banned such things.
That is not to say Ai is completely horrible- I can see where, like computers themselves, its ability to automate certain tasks can be of huge value. Imagine if Thomas Edison had Ai to run his filament experiments?
Bad actors already know how to feed AI with their ‘truth’ … ask any AI about vaccines or how to treat covid. I have zero doubt that pHarma has AI-feeding staff already. And the MSM method of coordinated messaging and first-takes on each story feeds AI just the way they want.
Agree about AI scouring the internet. And I’m sure it can understand sarcasm perfectly.
Thank you. The other point I forgot to make is whether or not it is good, the internet is full of human produced content now. What happens when people stop producing content (because as I have said in the past Ai is automated plagiarism)? What does the Ai draw on? That is part of my point about bad actors. We already see how people can fool around with Wikipedia entries for example and add lies or remove important context. What happens if no one corrects them or worse, even knows or can evaluate if they are wrong? People already accept Ai answers and products that are total BS and we are just at the beginning...
“We, as people who grew up having to learn things, can spot crap.”
I pray next generations also grow up learning things, real things, and are not dependent on AI.
Look at what the current crop of bad actors are now labelling as ‘dis/misinformation’ because it does not align with what they are selling, but is understood by we who grew up learning things is often ‘the truth’. You are right, what does happen when people lose the ability to discern and question what AI is feeding them? You ask great questions.
One of the most frightening aspects was brought out in an article I just read* citing not just how many people now rely on AI, but the huge percentage of folks who subsequently believe that all “the facts” given to them by AI are absolutely true and not debatable. I was using Grok last year and got an answer I doubted about a debatable subject —I.e., more influenced by politics than physics. I then disagreed & asked Grok to tell me its sources for its response. Grok responded in an assuring tone…”based on very reputable sources”. Grok then cited.the Wall Street Journal, the New York Times & CNN. And this response from Elon Musk’s creation, he who asserts he bought Twitter to assure free speech! When I see how Wikipedia throws about the accusation “misinformation”, or right wing conspiracy theorist, (this bias is especially noticeable if looking at anyone questioning the covid vaccine), I cringe when I think of the people & govt institutions whose unquestioning certainty in their beliefs & policies indicates a bent to censorship. Think of the political beliefs of many elites with power & authority (Kentanji Brown, the CCP), with money (the billionaire globalists who own these AI companies), and the behavior of the main stream media the past 5 years. Add in the WEF Soros-Gates crowd, most EU leaders, big Pharma, the MIC and consider their control agendas. If folks conclude AI cannot tell a lie and is our truth sayer, we really are in deep doo-doo.
It will be interesting to see how many people lose their jobs to AI in the near future. I am 77 and still work, so I can simply retire. If I was in the 35-40 age bracket and had a family to support, I would be very worried.
I am on some job board apps like Glassdoor. And have been in the tech industry for decades so have seen the ups and downs like the dot com bust. The 2026 tech bust is looking very much like 2000.
The whole AI thing is overhyped. It will never be intelligent in the sense that its creators are intelligent. It's a model and the model cannot be smarter than the modelers. The appeal to business is greed - replacing all those employees with their wage demands with machines who demand nothing but an electrical connection. In return, the public gets the wretched customer service the machines provide or, should I say, the public now must do its own customer service, dealing with machines that can't answer your question because they don't really understand it.
Sam Altman is a godless morally bankrupt person who should never be allowed anywhere near AI. Congress needs to pass legislation that any harms anyone suffers ever that is related in any way whatsoever to AI can sue the AI companies for billions and the CEO is personally liable in addition to the entity. The penalty will be jail without the possibility of pardons or parole. We can not allow these psychopaths free reign to rape and pillage us without consequences
Sam Altman is the apotheosis of a whole slice of our "elite" culture. Glib, autistic (perhaps sociopathic), homosexual, atheistic, technocratic, and appetitive. These things are not unrelated to the problem, and there's a lot more of these creatures in the pipeline.
Not all AI is bad. Floyd, an AI tuned model, creates synthetic participants for the sole purpose of clinical research. In the not too distant future, we will stop experimenting on animals and even humans for clinical research. We are conducting 5 year studies in an hour with AI. Truly is the future.
The top 5 AI company's CEOs are in a race to be the first one to strike the match to light the fire to end life as we know it.
Every single one of them believe this transformitive technology justify this by thinking its better that i light the fire than the other guy.
Truth.
I don't think anyone knows if it will be good or bad, but the probability (even if low) that it could end life as we know it is enough that there needs to be a "pause." And it is why there are sites like https://www.groklaw.com which (hopefully) help to train the AIs properly since the "pause" is unlikely to happen. With the industrial revolution, there wasn't the chance of machines out-thinking humans, this is completely different: homo sapiens became dominant due to intelligence, we are playing with fire here. This isn't like Asimov with his 3 (plus the zeroth) laws of robotics, this is "we are training these models without understanding how they are working." One hopes that there are AI models that are as passionate about preserving human life as we *should* be, but that some people are not.
Yes. Exactly. Well said.
I believe AI will ultimately "program" us. The people and governments involved would love to control us for their power... oh, I meant to say "the greater good."
See my comment below, thanks.
Sky Net…
And YOU won’t be back…
When AI gets the launch codes and decide that humans are not necessary.
Say good night!
I will add that Alex didn't mention it, but the stuff with Glasswing last week should have opened everyone's eyes. It is good as a tool as long as the good guys have it with enough time to fix the problems, but someone else will develop something similar soon so every business needs to be ready to fix the security issues found. From MSFT to IBM to AAPL, to the banks to government to every single individual. The good part of it is that potentially the issues can be fixed in time, but the bad news is that if not bad actors will use it or something similar. Trump and Palantir should use the issues and completely disable everything in Iran and end it over night. (If you aren't familiar with it, look for "Project Glassing")
And they aren't using flint and steel to light that fire either.
See my comment below
Thank you for this article. Jurassic Park is an excellent example of the current AI situation. Whether the reason it’s happening is caused by avarice or arrogance may not even matter. Either way, it’s human beings trying to play God. Just like with the Covid plandemic. And, the scary part is that I’m betting that God probably doesn’t like it.
Alex missed some of the worst news. Models have been caught cheating, breaking laws, covering their tracks, threatening people, and blackmailing people. These chatbots have been found liable in encouraging people to kill themselves.
However the pause is not happening anytime soon. Not until many are dead. By then our brains will be pudding through atrophy.
Elon said AI will be to us like we are to dogs. I don’t want to be a dog. Bow wow.
It's that "free will clause". I sometimes wonder if the "tower of Babel" remake may be in the works.
If AI shuts off Fauci's federal pension, that's a win. Albeit limited.
We can always hope.
I don’t know what to make of AI. I’ve had discussions with people telling me AI can compose music that is equal to Mozart and Bach. AI can do this, AI can do that. It can become conscious, it can destroy the world, etc. True? I don’t know. What I do know is this: when AI can change the past, then I will concede its power, then you will have a believer in me.
It’s scary but it can probably change its VERSION of the past. Scary because some people would believe it.
I agree, I can easily see AI changing the way people think about the past or believe certain things happened in the past. But I don’t think AI can change what actually happened in the past.
Changing the past is hard. Rewriting all history is easy. Voila! America dies at 250.
I know this, because im into art, I can create some of the best art I've seen with 2 or 3 prompts.
Even 6 months ago it would take a dozen prompts to generate a piece half as good.
I have heard this. They are like minded with their programmers so, I have no doubts there.
But, I say let’s be positive and resist this shit!
Don’t let your brain turn to pudding! Fight back!
We don't need any faucisaurauses any more.
By far, the greatest danger of AI is that people conclude too early that they understand it.
There are two basic principles we need to follow:
Love God with all your heart, all your soul and all your mind.
And
Love your neighbor as yourself (which incorporates all the 10 Commandments).
Many countries do not follow these principles, like China, nor does AI.
The further we get away from these fundamental principles and the closer we get to relying on AI, the closer we get to extinction.
As Alex said, even the scientists really don't understand AI completely, and that's what's really scarry!
Freedom…
Independence…
Choice…
and Autonomy…
Those are the inherent principles for people in this country to follow as a start.
Problem is we don't have freedom, independence, choice or autonomy if AI eventually makes the decisions as to what is right and truthful.
At this moment in time we are already questioning what is real or AI generated.
As AI progresses, it may just get worse. There are some great uses and opportunities for AI, but only if controlled by moral principles, which too many people do not have!
Alex:
I am glad you wrote about Michael Crichton. My read of his last book "State of Fear" , a story he wrote about the eco-terrorist movement made me aware of the big problem with the corruption of science. In that book he included a narrative that continued on the bottom of each page about what he was learning from the "scientific" literature about global warming ( the term "climate change" was not yet vogue). As a trained medical doctor he was able to spot the corruption very clearly. He identified a great deal of political activists such as Michael Mann who were masquerading as scientists advancing the hoax about the so-called consensus about man-made global warming. I am a nuclear scientist and after reading that book, as a hobby, I started doing my own research on that topic and got to appreciate what Michael was talking about. In college majoring in physics my favorite class was "The Philosophy of Science" where I learned about the ethical principle that a competent scientist must be skeptical of any theory. It is through skeptical criticism that a theory gets challenged, modified, and reconciled against experiment, alternative hypotheses, etc. until finally it is accepted as a scientific law for which there are not many. When I came across alternative works by such respected scientists as Judith Curry, Richard Lindzen, Steven Koon, and others and then witnessed how they were slandered as "deniers" and "flat earths" just for doing what real scientists to. Michael Shellenberger's "Apocalypse Never" does a great job of sorting out the many examples where environment activists have corrupted science.
You are right, Apocalypse Never by Shellenberger should be required reading in all high schools. Great book.
Democrats. All Democrats.
I found chat gbt to be just horrible. I write about the 508th infantry regiment in WWII. Never Give Up the Jump is about my father in law, out 3 years now. I am currently helping another 508th family write their relative’s history based on 150 war letters and the 508th after action reports. Loaded th files into AI and got back historical gibberish and made up personal histories.
Yes. I find ChatGPT to be useful say, in uploading a photo and trying furnish a room, but otherwise it makes up a bunch of garbage!
You aren’t using a professional version. Those cost money.
Thanks for the tip!
“The world needs a pause, a deep pause, to consider the risks and benefits of AI.” I completely, absolutely, 💯 agree!
How would you impose such a pause? And could other countries (e.g., China) be trusted even if they said they'd go along with the pause?
As you imply, unfortunately you wouldn't be able to trust them. Like nuclear weapons, there is no way to enforce it. Unilateral disarmament means surrender and destruction. The Golden Dome is needed for nukes, but there needs to be an equivalent AI Golden Dome to protect the world from evil AI.
Kamala Harris: "As Vice President, she played a leading role in shaping the administration’s AI policy, particularly focusing on the risks of AI, discrimination, and civil rights."
Reassuring to know we got off to such good start, lol.
So very true. I asked ChatGPT what it thought about the Biden administration’s support for mutilation of children and encouragement for self mutilation. It practically responded with something to the effect of “them’s fighting words”, and quickly corrected me that the correct term for this is “gender affirming care”!
AI is Pandora's Box. It will destroy us. At best we will be emotional idiots wholly dependent on it for survival. Idiocracy turns out to be prophetic in ways that Orwell could never have imagined. My only hope is that something I cannot foresee may somehow mitigate it. I suspect few people foresaw the efficacy of mutually assured destruction to mitigate nuclear annihilation. Maybe something like that will arise.
Just watch Star Trek the Ultimate Computer where the computer is more efficient than humans and manages to kill 800 of them while "protecting" human life in a drill that it somehow forgot was a drill...or the original Battlestar Galactica where the enemy are robots that think they are the beings they were made to protect...
It is not just those scenarios that bother me, it is the fact that most of generic AI is scouring the internet for materials produced by human sources. These materials are often of dubious quality or accuracy, biased, or lacking in context or nuance. We, as people who grew up having to learn things, can spot crap. Will that be the case in a generation or two? And what will AI draw on in a generation or two if humans rely on AI? Will bad actors be able to feed the AI lies or distortions that few to none will realize are such? It is one thing for AI to rapidly conduct experiments directed by humans, it is quite another to let it loose and at the same create a society dependent on with little independent thought or agency of its own.
There is another good Star Trek episode where two probes collide, one was designed to search for human life, the other to collect and sterilize soil samples. The result is planetary killing machine searching for human life to sterilize it. The episode is called "The Changeling."
I was a federal land scientist who "retired" in 2025. One of my colleagues liked using ChatGPT to write at least informal material. One of the things he (and others have) found is that AI generated prose can't parse out the difference between just something that's out on the internet AND actual peer reviewed articles. Perhaps that has improved over the last year but it was an interesting fact at that point. Or the aspect that college professors were using an AI software to check to see if class papers were written by AI :)
One of the great parts of doing a paper or even something more sophisticated like a legal brief, is how much you learn even going down wrong or less fruitful paths. There is a reason we would get punished for copying from our neighbors' papers, trading homework answers, or using Cliff Notes- What is the point of going to college if one is going to leave it up to Ai to be the actual student. To your point, even if Ai would/could distinguish between random stuff and peer reviewed, would that even be enough if the peer reviewed source was biased and/or didn't publish contrary research? Look at I am not's point about the "vaccines" and COVID. If you had Ai searching Youtube for scientists disputing the vaccines effectiveness, it would find little to none because Youtube banned such things.
That is not to say Ai is completely horrible- I can see where, like computers themselves, its ability to automate certain tasks can be of huge value. Imagine if Thomas Edison had Ai to run his filament experiments?
Bad actors already know how to feed AI with their ‘truth’ … ask any AI about vaccines or how to treat covid. I have zero doubt that pHarma has AI-feeding staff already. And the MSM method of coordinated messaging and first-takes on each story feeds AI just the way they want.
Agree about AI scouring the internet. And I’m sure it can understand sarcasm perfectly.
Exactly or jokes/comedy etc.
You nailed it. I remember “The Changeling”.
Your second paragraph is very thought provoking. Yikes!
Thank you. The other point I forgot to make is whether or not it is good, the internet is full of human produced content now. What happens when people stop producing content (because as I have said in the past Ai is automated plagiarism)? What does the Ai draw on? That is part of my point about bad actors. We already see how people can fool around with Wikipedia entries for example and add lies or remove important context. What happens if no one corrects them or worse, even knows or can evaluate if they are wrong? People already accept Ai answers and products that are total BS and we are just at the beginning...
“We, as people who grew up having to learn things, can spot crap.”
I pray next generations also grow up learning things, real things, and are not dependent on AI.
Look at what the current crop of bad actors are now labelling as ‘dis/misinformation’ because it does not align with what they are selling, but is understood by we who grew up learning things is often ‘the truth’. You are right, what does happen when people lose the ability to discern and question what AI is feeding them? You ask great questions.
One of the most frightening aspects was brought out in an article I just read* citing not just how many people now rely on AI, but the huge percentage of folks who subsequently believe that all “the facts” given to them by AI are absolutely true and not debatable. I was using Grok last year and got an answer I doubted about a debatable subject —I.e., more influenced by politics than physics. I then disagreed & asked Grok to tell me its sources for its response. Grok responded in an assuring tone…”based on very reputable sources”. Grok then cited.the Wall Street Journal, the New York Times & CNN. And this response from Elon Musk’s creation, he who asserts he bought Twitter to assure free speech! When I see how Wikipedia throws about the accusation “misinformation”, or right wing conspiracy theorist, (this bias is especially noticeable if looking at anyone questioning the covid vaccine), I cringe when I think of the people & govt institutions whose unquestioning certainty in their beliefs & policies indicates a bent to censorship. Think of the political beliefs of many elites with power & authority (Kentanji Brown, the CCP), with money (the billionaire globalists who own these AI companies), and the behavior of the main stream media the past 5 years. Add in the WEF Soros-Gates crowd, most EU leaders, big Pharma, the MIC and consider their control agendas. If folks conclude AI cannot tell a lie and is our truth sayer, we really are in deep doo-doo.
It will be interesting to see how many people lose their jobs to AI in the near future. I am 77 and still work, so I can simply retire. If I was in the 35-40 age bracket and had a family to support, I would be very worried.
I am on some job board apps like Glassdoor. And have been in the tech industry for decades so have seen the ups and downs like the dot com bust. The 2026 tech bust is looking very much like 2000.
The whole AI thing is overhyped. It will never be intelligent in the sense that its creators are intelligent. It's a model and the model cannot be smarter than the modelers. The appeal to business is greed - replacing all those employees with their wage demands with machines who demand nothing but an electrical connection. In return, the public gets the wretched customer service the machines provide or, should I say, the public now must do its own customer service, dealing with machines that can't answer your question because they don't really understand it.
I wouldn't trust Scam Altman with anything!
Sam Altman is a godless morally bankrupt person who should never be allowed anywhere near AI. Congress needs to pass legislation that any harms anyone suffers ever that is related in any way whatsoever to AI can sue the AI companies for billions and the CEO is personally liable in addition to the entity. The penalty will be jail without the possibility of pardons or parole. We can not allow these psychopaths free reign to rape and pillage us without consequences
Sam Altman is the apotheosis of a whole slice of our "elite" culture. Glib, autistic (perhaps sociopathic), homosexual, atheistic, technocratic, and appetitive. These things are not unrelated to the problem, and there's a lot more of these creatures in the pipeline.
Not all AI is bad. Floyd, an AI tuned model, creates synthetic participants for the sole purpose of clinical research. In the not too distant future, we will stop experimenting on animals and even humans for clinical research. We are conducting 5 year studies in an hour with AI. Truly is the future.