There is a lot of speculation and unverified facts (bullshit) in this article. If the government confirmed that Altman was conspiring illegally with the Chinese to gather more information for his models, why wasn’t he arrested instead of just being fired? And why was he rehired? If the Chinese were investing in his efforts to create an AGI why would they tip off the Biden Administration that he was conspiring with their military? Why would they tell Biden anything. If D2 was working with Altman it would have been approved at the highest levels of the Chinese Communist Party, that is if D2 even is a part of China’s intelligence operations. Ms. Malmgren’s story doesn’t pass the smell test.
The story that has been reported that last week’s drama was a dust up between factions on the OpenAI board over commercialization versus safety with Altman’s faction winning is probably the truth. But that doesn’t satisfy people who would rather spin wild tales of international intrigue and AI causing our imminent destruction. Ms. Malmgren should be writing for the movies. Maybe she already is, but that’s just speculation on my part.
Agreed Benjamin, this article is clickbait and not worth the read. Sad to see Unherd drop it’s quality metrics so low here. Honest coverage of the OpenAI board and their unhealthy obsession with Safetyism is the story. The bigger danger is overzealous regulators using narratives from hysterics like Dr. Malmgren as excuse for limiting AGI to a few major players who will work with the Administrative State to control the market and public narratives to serve their limited interests. We saw this with Big Pharma and their undue influence on the American people through their capture of the Administrative State and the public narrative throughout COVID(see Missouri v Biden). Free markets unfettered by self interested bureaucrats, allowing innovation from the competition of a diversity of players, is the safest path forward, not centralized authority over this emerging technology.
“If the government confirmed that Altman was conspiring illegally with the Chinese to gather more information for his models, why wasn’t he arrested instead of just being fired?”
The point of the article is that we’ve run headlong into a situation where even if we think there isn’t a national security interest at stake, we’re not even sure if it isn’t an international security issue at stake.
Last edited 5 months ago by John Riordan
Matt Sylvestre
5 months ago
The West must realize that what ever “guard rails” we try to impose on AI, The East (China) will not necessarily respect. China is driven by its own lights, its own self interest… This stuff is happening no matter if we try to slow it down or not for better or for worse…
While I admire your optimism regarding “global discussion” the stakes are so high that talking will have no effect on the development of these technologies.
Power–economic and military–as well as the sheer, horrific, potential to control all aspects of the individual lives of people everywhere are the potential goals.
The counter to that (the argument Max Tegmark and Geoffrey Hinton make) is that we do have observed worldwide bans on technologies generally seen to be threatening to humanity – human cloning, biological weapons etc
I remain convinced that although AI may well be a civilisational threat, the form this threat takes is in the use by hostile states of AI in modern warfare, both of the cold and hot variety. I do not think that humans as a whole will find themselves in a war with AI.
AI is a weapon, not a competing civilisation where the winner will be decided by technological supremacy.
I agree. So much has been made of the existential danger of the very existence of AI that very little heed has been paid to how AI might and likely will be weaponized. In that respect, AI is yet another in an ever expanding array of weapons humans can use to fight, harm, exploit, and kill one another for power or profit.
Christopher Barclay
5 months ago
If Altman thinks that the Chinese don’t believe in borders or nationality, he is very naive.
This law reflects an intuition I felt a lot over the last few years of reading question headlines.
Thank you for the link.
Neil Chapman
5 months ago
Not sure most folks really appreciate the speed of progress and development in the AI sector. Mustafa Suleyman’ s book “The Coming Wave” is a good place to start.
There’s a hell of a bandwagon, for sure. But the US government wouldn’t need to be so circumspect if it genuinely felt Sam was imperilling national security. And why would Xi tell Biden about Altman Chinese connections? This whole piece is a bit of a stretch.
And would Biden understand what he had been told? Evidence suggests no.
Alex Colchester
5 months ago
Humans are very poor at predicting things. In the 1950’s it was assumed that a computer would never play chess better than a grandmaster but we would all be served in our houses by robotic butlers. Solving Chess was easy but it’s still impossible to get a robot to cook you dinner. Whatever everyone thinks AI is going to do, it probably will fail at. What we need to focus on is what we think it won’t be able to do.
If you think the vast and massively funded military industrial complex has not for decades been trying (and failing) to get a robot to look in your fridge and prepare dinner (because if it can solve that randomly presented spatial riddle, then it can also easily break into your house and kill you) then you are confidently deluded.
I like to think of it as a ‘deluded confident impossible’
Similar in the vein to the old belief that a human could never travel faster than a galloping horse
You’ve completely missed my point. It was in the 1950’s that everyone was convinced they were imminent. Your implication they are 3 years away means they have come 76 years after they were ‘just around the corner’.
iambic mouth
5 months ago
There are many claims in this article tgat aren’t true. Quantum simulator, not even quantum computer can perform calculations at this speed, and these simulators are all over the world. To get diverse data samples we need to use Chinese or Far East data? Does the author know how Chat GPT was even trained? And, the white male hypothesis in clinical trial data has been disproven times and times again.
Agreed 100%. The “white make tech bros” straw man is a useless and fictional identity politics construct. But even if it were true, does she suppose that these guys would be too stupid to understand this and develop a workaround?
We should also question her premise about AI and “true diversity,” lest we once again drift into fuzzy-headed woke slop. DOES and SHOULD the world want “AI designed for true diversity”? DOES and SHOULD it want “medical and financial products created to suit the broadest range of humans”? Is that really the best path forward? Maybe so, but it’s worth asking.
Last edited 5 months ago by UnHerd Reader
Peter Principle
5 months ago
The “threat” of AI is not that it can out-smart us, but rather that we have been bamboozled into believing that it is smarter than it really is, so we become over-reliant on it.
Most Unherd readers will have experience of colleagues who sound highly knowledgeable and convincing (especially to the boss), yet who are unreliable. Well, that is my impression of the technology. When it makes a mistake and you correct it, it memorises your correction, but it is not so good at making appropriate changes to the reasoning that led to the mistake.
no, it is an existential threat – at the lease will destroy family and values.
Saul D
5 months ago
The Q* (Q-Star) work is coming in hot from what I read, as it seems to accelerate problem solving and enhance what looks like conceptual thinking. So, for instance, ChatGPT sounds fabulous but actually isn’t that great at consistent logic or arithmetic without help (think an English Language Student opining on an engineering problem).
However, as I understand, Q* shortcuts some type of model generation processes and this seems to boost conceptual understanding, so it ‘gets’ logic and maths better and faster, and this is likely to improve over time.
Consequently, AI has the real potential of getting smarter than us and what we know now, with the potential of pushing beyond our current knowledge boundaries.
The challenge is that AI may become the equivalent of an autistic savant – huge intellectual skills, but struggling with empathy and human-sensitivities. If AI has power or influence (event through recommendation systems) that could put people at risk. What seems logical is not always right.
However, for us, like Pandora, this is not a box we can close. Our skills at spotting, judging and interacting with AI-generated content are only just beginning, and our skepticism about received opinion needs to go up another couple of orders of magnitude.
Andrew Thompson
5 months ago
Yes let’s all ban all research into this dangerous thing until we understand it a bit better. Okay, great idea we won’t research if you promise not to research. Okay then deal….(Research all round it is then guys)
Felix Hornoiu
5 months ago
We will look back on this and laugh hard. I find this a tech version of Joe Rogans MMA episodes, lots of gossip and speculations over facts that can’t be verified.
Whenever there’s a rumble into some (corporation) issue it’s easy just to follow the money. That’s it!
Last edited 5 months ago by Felix Hornoiu
Kolya Wolf
4 months ago
More intelligence is a Good Thing.
Douglas Redmayne
5 months ago
Hopefully the American middle class tech bros of the US will succeed in generating AGI first and creating its vase code otherwise a Woke or a Chinese AGI will be bad news.
Just for fun, some Poetry, and Biblical reference to this approaching horror:The Second Coming By W B Yeats Turning and turning in the widening gyre The falcon cannot hear the falconer; Things fall apart; the centre cannot hold; Mere anarchy is loosed upon the world, The blood-dimmed tide is loosed, and everywhere The ceremony of innocence is drowned; The best lack all conviction, while the worst Are full of passionate intensity.
Surely some revelation is at hand; Surely the Second Coming is at hand. The Second Coming! Hardly are those words out When a vast image out of Spiritus Mundi Troubles my sight: somewhere in sands of the desert A shape with lion body and the head of a man, A gaze blank and pitiless as the sun, Is moving its slow thighs, while all about it Reel shadows of the indignant desert birds. The darkness drops again; but now I know That twenty centuries of stony sleep Were vexed to nightmare by a rocking cradle, And what rough beast, its hour come round at last, Slouches towards Bethlehem to be born?
n/a
Source: The Collected Poems of W. B. Yeats (1989)
KJB Revelations 13
”The Beast Out of the Earth 11And I beheld another beast coming up out of the earth; and he had two horns like a lamb, and he spake as a dragon. 12And he exerciseth all the power of the first beast before him, and causeth the earth and them which dwell therein to worship the first beast, whose deadly wound was healed. 13And he doeth great wonders, so that he maketh fire come down from heaven on the earth in the sight of men, 14And deceiveth them that dwell on the earth by the means of those miracles which he had power to do in the sight of the beast; saying to them that dwell on the earth, that they should make an image to the beast, which had the wound by a sword, and did live. 15And he had power to give life unto the image of the beast, that the image of the beast should both speak, and cause that as many as would not worship the image of the beast should be killed.
The Mark of the Beast 16And he causeth all, both small and great, rich and poor, free and bond, to receive a mark in their right hand, or in their foreheads: 17And that no man might buy or sell, save he that had the mark, or the name of the beast, or the number of his name. 18Here is wisdom. Let him that hath understanding count the number of the beast: for it is the number of a man; and his number is Six hundred threescore and six.’;;
There is a lot of speculation and unverified facts (bullshit) in this article. If the government confirmed that Altman was conspiring illegally with the Chinese to gather more information for his models, why wasn’t he arrested instead of just being fired? And why was he rehired? If the Chinese were investing in his efforts to create an AGI why would they tip off the Biden Administration that he was conspiring with their military? Why would they tell Biden anything. If D2 was working with Altman it would have been approved at the highest levels of the Chinese Communist Party, that is if D2 even is a part of China’s intelligence operations. Ms. Malmgren’s story doesn’t pass the smell test.
The story that has been reported that last week’s drama was a dust up between factions on the OpenAI board over commercialization versus safety with Altman’s faction winning is probably the truth. But that doesn’t satisfy people who would rather spin wild tales of international intrigue and AI causing our imminent destruction. Ms. Malmgren should be writing for the movies. Maybe she already is, but that’s just speculation on my part.
Agreed Benjamin, this article is clickbait and not worth the read. Sad to see Unherd drop it’s quality metrics so low here. Honest coverage of the OpenAI board and their unhealthy obsession with Safetyism is the story. The bigger danger is overzealous regulators using narratives from hysterics like Dr. Malmgren as excuse for limiting AGI to a few major players who will work with the Administrative State to control the market and public narratives to serve their limited interests. We saw this with Big Pharma and their undue influence on the American people through their capture of the Administrative State and the public narrative throughout COVID(see Missouri v Biden). Free markets unfettered by self interested bureaucrats, allowing innovation from the competition of a diversity of players, is the safest path forward, not centralized authority over this emerging technology.
“If the government confirmed that Altman was conspiring illegally with the Chinese to gather more information for his models, why wasn’t he arrested instead of just being fired?”
The point of the article is that we’ve run headlong into a situation where even if we think there isn’t a national security interest at stake, we’re not even sure if it isn’t an international security issue at stake.
The West must realize that what ever “guard rails” we try to impose on AI, The East (China) will not necessarily respect. China is driven by its own lights, its own self interest… This stuff is happening no matter if we try to slow it down or not for better or for worse…
Yes. Seems like this is a passage humanity has to go through; an event that will demand a global discussion at some point in time.
While I admire your optimism regarding “global discussion” the stakes are so high that talking will have no effect on the development of these technologies.
Power–economic and military–as well as the sheer, horrific, potential to control all aspects of the individual lives of people everywhere are the potential goals.
The counter to that (the argument Max Tegmark and Geoffrey Hinton make) is that we do have observed worldwide bans on technologies generally seen to be threatening to humanity – human cloning, biological weapons etc
Do observed bans mean that it doesn’t happen, or that the research isn’t published?
Do we really know that biological weapons aren’t being developed?
Very naive take on the world.
I remain convinced that although AI may well be a civilisational threat, the form this threat takes is in the use by hostile states of AI in modern warfare, both of the cold and hot variety. I do not think that humans as a whole will find themselves in a war with AI.
AI is a weapon, not a competing civilisation where the winner will be decided by technological supremacy.
The good news is that we have David Cameron with his science background back as Foreign Secretary.
I agree. He’s going to save us all for sure.
Ha! Ha!
I agree. So much has been made of the existential danger of the very existence of AI that very little heed has been paid to how AI might and likely will be weaponized. In that respect, AI is yet another in an ever expanding array of weapons humans can use to fight, harm, exploit, and kill one another for power or profit.
If Altman thinks that the Chinese don’t believe in borders or nationality, he is very naive.
The Chinese believe in their own borders, not sure about everyone elses though.
https://en.m.wikipedia.org/wiki/Betteridge%27s_law_of_headlines
That was interesting! Thanks!
This law reflects an intuition I felt a lot over the last few years of reading question headlines.
Thank you for the link.
Not sure most folks really appreciate the speed of progress and development in the AI sector. Mustafa Suleyman’ s book “The Coming Wave” is a good place to start.
There’s a hell of a bandwagon, for sure. But the US government wouldn’t need to be so circumspect if it genuinely felt Sam was imperilling national security. And why would Xi tell Biden about Altman Chinese connections? This whole piece is a bit of a stretch.
And would Biden understand what he had been told? Evidence suggests no.
Humans are very poor at predicting things. In the 1950’s it was assumed that a computer would never play chess better than a grandmaster but we would all be served in our houses by robotic butlers. Solving Chess was easy but it’s still impossible to get a robot to cook you dinner. Whatever everyone thinks AI is going to do, it probably will fail at. What we need to focus on is what we think it won’t be able to do.
I’m mildy optimistic that a robot will cook my great-great-great-great-granddaughter’s dinner.
I’m sure a robot could be built to cook a dinner but that’s not where the makers are focused.
If you think the vast and massively funded military industrial complex has not for decades been trying (and failing) to get a robot to look in your fridge and prepare dinner (because if it can solve that randomly presented spatial riddle, then it can also easily break into your house and kill you) then you are confidently deluded.
Would that be a known-unknown, or one of those unknown-unknowns ?
I like to think of it as a ‘deluded confident impossible’
Similar in the vein to the old belief that a human could never travel faster than a galloping horse
You need to keep up. Robot servants are very likely once AGI arrives. This is expected within 3 years.
You’ve completely missed my point. It was in the 1950’s that everyone was convinced they were imminent. Your implication they are 3 years away means they have come 76 years after they were ‘just around the corner’.
There are many claims in this article tgat aren’t true. Quantum simulator, not even quantum computer can perform calculations at this speed, and these simulators are all over the world. To get diverse data samples we need to use Chinese or Far East data? Does the author know how Chat GPT was even trained? And, the white male hypothesis in clinical trial data has been disproven times and times again.
Agreed 100%. The “white make tech bros” straw man is a useless and fictional identity politics construct. But even if it were true, does she suppose that these guys would be too stupid to understand this and develop a workaround?
We should also question her premise about AI and “true diversity,” lest we once again drift into fuzzy-headed woke slop. DOES and SHOULD the world want “AI designed for true diversity”? DOES and SHOULD it want “medical and financial products created to suit the broadest range of humans”? Is that really the best path forward? Maybe so, but it’s worth asking.
The “threat” of AI is not that it can out-smart us, but rather that we have been bamboozled into believing that it is smarter than it really is, so we become over-reliant on it.
Most Unherd readers will have experience of colleagues who sound highly knowledgeable and convincing (especially to the boss), yet who are unreliable. Well, that is my impression of the technology. When it makes a mistake and you correct it, it memorises your correction, but it is not so good at making appropriate changes to the reasoning that led to the mistake.
no, it is an existential threat – at the lease will destroy family and values.
The Q* (Q-Star) work is coming in hot from what I read, as it seems to accelerate problem solving and enhance what looks like conceptual thinking. So, for instance, ChatGPT sounds fabulous but actually isn’t that great at consistent logic or arithmetic without help (think an English Language Student opining on an engineering problem).
However, as I understand, Q* shortcuts some type of model generation processes and this seems to boost conceptual understanding, so it ‘gets’ logic and maths better and faster, and this is likely to improve over time.
Consequently, AI has the real potential of getting smarter than us and what we know now, with the potential of pushing beyond our current knowledge boundaries.
The challenge is that AI may become the equivalent of an autistic savant – huge intellectual skills, but struggling with empathy and human-sensitivities. If AI has power or influence (event through recommendation systems) that could put people at risk. What seems logical is not always right.
However, for us, like Pandora, this is not a box we can close. Our skills at spotting, judging and interacting with AI-generated content are only just beginning, and our skepticism about received opinion needs to go up another couple of orders of magnitude.
Yes let’s all ban all research into this dangerous thing until we understand it a bit better. Okay, great idea we won’t research if you promise not to research. Okay then deal….(Research all round it is then guys)
We will look back on this and laugh hard. I find this a tech version of Joe Rogans MMA episodes, lots of gossip and speculations over facts that can’t be verified.
Whenever there’s a rumble into some (corporation) issue it’s easy just to follow the money. That’s it!
More intelligence is a Good Thing.
Hopefully the American middle class tech bros of the US will succeed in generating AGI first and creating its vase code otherwise a Woke or a Chinese AGI will be bad news.
Just for fun, some Poetry, and Biblical reference to this approaching horror:The Second Coming By W B Yeats
Turning and turning in the widening gyre
The falcon cannot hear the falconer;
Things fall apart; the centre cannot hold;
Mere anarchy is loosed upon the world,
The blood-dimmed tide is loosed, and everywhere
The ceremony of innocence is drowned;
The best lack all conviction, while the worst
Are full of passionate intensity.
Surely some revelation is at hand;
Surely the Second Coming is at hand.
The Second Coming! Hardly are those words out
When a vast image out of Spiritus Mundi
Troubles my sight: somewhere in sands of the desert
A shape with lion body and the head of a man,
A gaze blank and pitiless as the sun,
Is moving its slow thighs, while all about it
Reel shadows of the indignant desert birds.
The darkness drops again; but now I know
That twenty centuries of stony sleep
Were vexed to nightmare by a rocking cradle,
And what rough beast, its hour come round at last,
Slouches towards Bethlehem to be born?
n/a
Source: The Collected Poems of W. B. Yeats (1989)
KJB Revelations 13
”The Beast Out of the Earth
11And I beheld another beast coming up out of the earth; and he had two horns like a lamb, and he spake as a dragon. 12And he exerciseth all the power of the first beast before him, and causeth the earth and them which dwell therein to worship the first beast, whose deadly wound was healed. 13And he doeth great wonders, so that he maketh fire come down from heaven on the earth in the sight of men, 14And deceiveth them that dwell on the earth by the means of those miracles which he had power to do in the sight of the beast; saying to them that dwell on the earth, that they should make an image to the beast, which had the wound by a sword, and did live. 15And he had power to give life unto the image of the beast, that the image of the beast should both speak, and cause that as many as would not worship the image of the beast should be killed.
The Mark of the Beast
16And he causeth all, both small and great, rich and poor, free and bond, to receive a mark in their right hand, or in their foreheads: 17And that no man might buy or sell, save he that had the mark, or the name of the beast, or the number of his name. 18Here is wisdom. Let him that hath understanding count the number of the beast: for it is the number of a man; and his number is Six hundred threescore and six.’;;
Altman….