With his dismal track record, that anyone still listens to Elon's predictions boggles the mind. That said, there are a lot of things about our current era that boggle the mind.
The FSD bloopers are stupefying. Missy Cummings tears it to shreds on LinkedIn continuously.
Having worked on projects involved SEMI-automated TRAINS, I'm here to tell you that even they are really hard to auto-drive. And they don't even have to steer.
Most metro airports have some flavor of people mover. Never noticed a driver in those, maybe because those systems were built to be automated instead of self driving being a bolt-on feature for an existing environment.
Maybe the trains thing would benefit from the kind of engineering talent he recruits.
Usually the way, because rail networks can be built up from a bare footprint (see Riyadh, Sydney Metro). For existing networks it's much more of a challenge. However, as I understand it, Musk's group are wedded to a visual/LIDAR approach - i.e. each individual car does everything. Automated trains, on the other hand, talk to the track and to the other trains.
In a way, FSD is a victim of a kind of self-fulfilling prophecy (not helped by Elon Musk continually making himself as popular as a sarin gas leak). There isn't enough demand for FSD, so there isn't the impetus to develop infrastructure to support FSD cars because the cost-benefit is lousy, which means that FSD has to rely on the crappiest possible approach, which means nobody wants FSD, which means there isn't enough demand for FSD...
The quest for driverless will probably influence interstate highway updates. I can see an interim result where driverless tractor trailer rigs pass between terminals at city edges where their loads are broken down and transferred to manually driven or driver assist vehicles for last mile delivery. This would be a bit like oceangoing vessels that are met by pilots at the harbor entrance and guided outbound similarly but once clear, run on their own.
Waymo-type urban taxis and self driving cars for individuals seem less appealing both economically and politically because of risk exposure.
It's amazing how he appears to think about it when he's asked the question, like he's thinking "wow, here's a question about autonomous driving that's totally new and I could not have expected; I better pause for a moment and calculate this thing that I've never thought about before."
Sally: I cannot help but wonder where the slapstick, "Airplane" like comedies are. My imagination goes back to Laugh-In or Seinfeld, or WKRP in Cincinnati, or movies like Young Frankenstein. Is it so serious that we have we lost our sense of humor?
Sally, yes, everything is upside down. My God, Trump is even rewarding the people who attacked the U.S. Capitol--after letting them all out of jail, of all things. (Poor things, what's wrong, after all, with attacking the Capitol Police, not to mention the seat of democracy?)
He's recruited some really talented people and set them onto really challenging tasks while exhibiting remarkable patience with failures along the way. He's made a business model to fit this. Who would have predicted in, say, 2000 that he'd put NASA on the ropes or provide global wireless data connectivity. Edison's first bulb and recording device were later refined to actually be reliable and utilitarian. In contrast, Musk seems to take his quests all the way to common reality. Yeah, he's a blowhard but who wouldn't be with his track record?
Indeed. Most of the fault lies with the mainstream media that just "parrot" the rumors they get without even questioning ... And then, there are the gullible people in always increasing number ("it must be true, I read it on Internet ...")
It'll be a bad morning for the pig who gets operated on by a test Optimus surgeon.
I still see the key-jangling that we saw with the Anthropic hacks and the maths solutions. Increasingly (and at the risk of sounding like Alex Jones), the mainstream media is wising up to the limitations and risks of the technology. They've also noticed the bubble. Hence, Elon et al need something, anything, to get the show back on track.
This is such an obvious Muskian response to his tanking Space X stock prices, which are set to plummet tomorrow as well. It’s like clockwork, every time his company is seeing bad news in the press, he invents a new Shiny Object to attempt to pump the stock price and deflect the narrative. His track record for delivering on these far-fetched promises speaks for itself. Don’t fall for the distraction, not worth arguing whether or not it’s possible. That’s not the point.
It’s like the whole point of DOGE was to steal our government records, not to save any taxpayer money. All those arguing about whether he succeeded or failed to cut government spending completely miss the point. It was a smokescreen all along.
By 2029? I’m a spine surgeon. How much can I bet? I’ll bet as much as possible. That mathematically makes no sense. The amount of possible chess moves on a chess board has only been estimated around 10^120 and has not been exactly calculated. There are more variables to an exponential degree in surgery.
Yesterday on the SpaceX earnings call he said they would put 1 million tons of cargo into orbit - which would require more than 27 starship launches per day.
Because he can’t help himself, he said 10 million was also possible.
Not as harmful as claiming one of his robots will perform surgery, but just as absurd.
His wildest medical claim remains, imho, the claim that his neuralink implants would make quadriplegics walk again
If your doctor uses AI and you've tried to schedule an appointment using their AI system, you've probably gotten very frustrated. It's ridiculously hilarious to think that something that can't schedule an appointment in 2026 is going to do cardiac procedures by 2030.
And help me here - what is Elon's expertise with medicine?
I agree that Musk’s claim about Optimus becoming better than the best human surgeons by 2029 is preposterous. But looks like two separate issues are involved.
First, “surgery” is not one uniform task. A purpose-built robotic system may conceivably become as good as a competent human surgeon in a small number of highly standardized and relatively simple procedures by the end of the decade. My guess is that these would represent no more than 5–10% of surgical cases. That is very different from being better than the best surgeons across surgery generally. The most difficult and variable cases (the most complex 1/3?) will probably remain beyond autonomous robots for another 2-3 decades, and possibly longer.
Second, surgical expertise contains a large amount of tacit knowledge. Skilled surgeons do not necessarily have conscious access to, or the ability to formalize, every judgment and adjustment they make during an operation. In one study I discuss in Chapter 16 of my book (Artificial Intelligence: Utopia, Dystopia, or something in between?), experienced surgeons could articulate in advance only ~30% of the cognitive steps they subsequently performed. The rest emerged only through detailed review, and some of it was not consciously accessible even to the surgeons themselves.
This does not mean that robots cannot learn such skills, but it does mean that knowledge cannot simply be extracted from surgeons, written down, and transferred into an AI system. Progress is much more likely to come through narrow, procedure-specific systems working with human surgeons than through a general-purpose humanoid robot replacing them.
People like Musk repeatedly underestimate the complexity of human judgment and experience, as well as the complexity and unpredictability of the real world. The best application of AI in surgery is going to be as an enhancement of human surgeons' ability. For example, as an overwatch that follows the surgeon's work, calls out issues or anomalies, and reminds the surgeon if a step is being missed or an optional task is recommended. It can also enhance the precision with which the surgeon operates (as is already done with machinery or robotics in many surgery applications).
I agree. AI’s most valuable near-term role will be to enhance the surgeon’s judgment, situational awareness, and precision rather than replace the surgeon.
There may be exceptional situations, however, in which autonomous AI-guided surgery becomes justified: for example, in Antarctica, during spaceflight, eventually on Mars, where no qualified surgeon is available and delay could be dangerous. Even then, I would expect this to be limited to relatively simple, highly standardized, and urgent procedures, ideally with remote human supervision whenever possible.
In ordinary clinical settings, and especially in complex or unpredictable operations, a human surgeon will remain involved for many years.
Elon Musk is just an embarrassing human. I have no idea how much of his wealth is "deserved", I have no idea how shrewd of a business man he is, I have no idea if he's an elite gamer like he claims (seems about as credible as the claim above), but I do know he's a deeply embarrassing person. Guy needs to stop digging.
He's from a very wealthy family. He started off on a low rung in Silicon Valley and, to his credit, has built up from there and learned from early mistakes. He did start off in a very good position (particularly compared to the median South African).
>Elite gamer
Certainly not. There's scores of YouTube videos showing the direct opposite.
With his dismal track record, that anyone still listens to Elon's predictions boggles the mind. That said, there are a lot of things about our current era that boggle the mind.
hard to know where to start
I love this supercut of his FSD cars.
https://www.youtube.com/watch?v=B4rdISpXigM
The FSD bloopers are stupefying. Missy Cummings tears it to shreds on LinkedIn continuously.
Having worked on projects involved SEMI-automated TRAINS, I'm here to tell you that even they are really hard to auto-drive. And they don't even have to steer.
Most metro airports have some flavor of people mover. Never noticed a driver in those, maybe because those systems were built to be automated instead of self driving being a bolt-on feature for an existing environment.
Maybe the trains thing would benefit from the kind of engineering talent he recruits.
Usually the way, because rail networks can be built up from a bare footprint (see Riyadh, Sydney Metro). For existing networks it's much more of a challenge. However, as I understand it, Musk's group are wedded to a visual/LIDAR approach - i.e. each individual car does everything. Automated trains, on the other hand, talk to the track and to the other trains.
In a way, FSD is a victim of a kind of self-fulfilling prophecy (not helped by Elon Musk continually making himself as popular as a sarin gas leak). There isn't enough demand for FSD, so there isn't the impetus to develop infrastructure to support FSD cars because the cost-benefit is lousy, which means that FSD has to rely on the crappiest possible approach, which means nobody wants FSD, which means there isn't enough demand for FSD...
"However, as I understand it, Musk's group are wedded to a visual/LIDAR approach"
They're wedded to visual ONLY at Musk's direction. No LIDAR, RADAR, etc. So even worse.
Well there you go. Yes, that's much worse.
You know that'll make our self-driving car better? Not giving it depth perception!
The quest for driverless will probably influence interstate highway updates. I can see an interim result where driverless tractor trailer rigs pass between terminals at city edges where their loads are broken down and transferred to manually driven or driver assist vehicles for last mile delivery. This would be a bit like oceangoing vessels that are met by pilots at the harbor entrance and guided outbound similarly but once clear, run on their own.
Waymo-type urban taxis and self driving cars for individuals seem less appealing both economically and politically because of risk exposure.
It's amazing how he appears to think about it when he's asked the question, like he's thinking "wow, here's a question about autonomous driving that's totally new and I could not have expected; I better pause for a moment and calculate this thing that I've never thought about before."
When there are enough failed predictions to fill whole websites and Wikipedia Pages, that's a strong sign that you should reconsider before you speak
Sally: I cannot help but wonder where the slapstick, "Airplane" like comedies are. My imagination goes back to Laugh-In or Seinfeld, or WKRP in Cincinnati, or movies like Young Frankenstein. Is it so serious that we have we lost our sense of humor?
It's more like reality turned into slapstick
Sally, yes, everything is upside down. My God, Trump is even rewarding the people who attacked the U.S. Capitol--after letting them all out of jail, of all things. (Poor things, what's wrong, after all, with attacking the Capitol Police, not to mention the seat of democracy?)
Trump is disgustingly blatant in his corruption
He's recruited some really talented people and set them onto really challenging tasks while exhibiting remarkable patience with failures along the way. He's made a business model to fit this. Who would have predicted in, say, 2000 that he'd put NASA on the ropes or provide global wireless data connectivity. Edison's first bulb and recording device were later refined to actually be reliable and utilitarian. In contrast, Musk seems to take his quests all the way to common reality. Yeah, he's a blowhard but who wouldn't be with his track record?
Indeed. Most of the fault lies with the mainstream media that just "parrot" the rumors they get without even questioning ... And then, there are the gullible people in always increasing number ("it must be true, I read it on Internet ...")
The mainstream media are stochastic parrots, which is why they are so easily replaced by LLMs
Both China and Russia are doing Starlink like projects!
While aside the NASA regulars he has more competition for orbit.
Off by a decade? In this instance, he's quite possibly off by a century.
Besides musk being off a decade and being off a century he is also OFF HIS FOCKING MIND🍀💙✌🏽
Yes, absurd. He's become so obsessed with his fantasy of AI and robots being better than humans. What is it about people he hates so much?
He'll never take your bet.
"What is it about people he hates so much?" Maybe he looked in the mirror one too many times.
The bet should be for Elon to be the first human to be operated on in 2029 😁
Can we do moving operations in driverless cars? (If he wants to save lives, what about SAiD?)
It'll be a bad morning for the pig who gets operated on by a test Optimus surgeon.
I still see the key-jangling that we saw with the Anthropic hacks and the maths solutions. Increasingly (and at the risk of sounding like Alex Jones), the mainstream media is wising up to the limitations and risks of the technology. They've also noticed the bubble. Hence, Elon et al need something, anything, to get the show back on track.
The Greatest Show on Earth (aka a circus)
This is such an obvious Muskian response to his tanking Space X stock prices, which are set to plummet tomorrow as well. It’s like clockwork, every time his company is seeing bad news in the press, he invents a new Shiny Object to attempt to pump the stock price and deflect the narrative. His track record for delivering on these far-fetched promises speaks for itself. Don’t fall for the distraction, not worth arguing whether or not it’s possible. That’s not the point.
It’s like the whole point of DOGE was to steal our government records, not to save any taxpayer money. All those arguing about whether he succeeded or failed to cut government spending completely miss the point. It was a smokescreen all along.
Agree. Distraction, distraction, distraction.
By 2029? I’m a spine surgeon. How much can I bet? I’ll bet as much as possible. That mathematically makes no sense. The amount of possible chess moves on a chess board has only been estimated around 10^120 and has not been exactly calculated. There are more variables to an exponential degree in surgery.
LLM do image matching, word matching and best you get is inference.
These snake oil salesmen have not consider the philosophies of logic and knowledge.
They are preachers hoping for faith...
I suppose the most important thing to say about Musk is why anybody takes anything this racist buffoon says seriously or even cares?
Yesterday on the SpaceX earnings call he said they would put 1 million tons of cargo into orbit - which would require more than 27 starship launches per day.
Because he can’t help himself, he said 10 million was also possible.
Not as harmful as claiming one of his robots will perform surgery, but just as absurd.
His wildest medical claim remains, imho, the claim that his neuralink implants would make quadriplegics walk again
and that's if Starship performed as advertized ... which it doesn't
Elon’s big claim about Starship (the one that everything hinges on) is that it’s reusable,which it isn’t.
It’s not even usable once, at least not by humans who want to be reusable after flying on it.
LOL.
Yeah, I have a piece in my queue about "reusability". Don't forget he said "rapidly reusable" - like same day. Never. Going. To. Happen.
Can't wait for Flight 14 - I really hope they try to "catch" the Starship on the pad. Rest of might get some more information.
Assuming Starship performs as advertised gets you to absurd conclusions, and if it falls even slightly short, the whole thing falls apart
Elon still pumping Starship like it’s going to change the world
https://thestandarddispatchandmail.substack.com/p/a-primer-on-elon-musk-speak?r=hllyh&utm_medium=ios
My rule of thumb is that if Elon Musk says something will happen, automatically assume it won't
He’s, if nothing, consistent!
If your doctor uses AI and you've tried to schedule an appointment using their AI system, you've probably gotten very frustrated. It's ridiculously hilarious to think that something that can't schedule an appointment in 2026 is going to do cardiac procedures by 2030.
And help me here - what is Elon's expertise with medicine?
I agree that Musk’s claim about Optimus becoming better than the best human surgeons by 2029 is preposterous. But looks like two separate issues are involved.
First, “surgery” is not one uniform task. A purpose-built robotic system may conceivably become as good as a competent human surgeon in a small number of highly standardized and relatively simple procedures by the end of the decade. My guess is that these would represent no more than 5–10% of surgical cases. That is very different from being better than the best surgeons across surgery generally. The most difficult and variable cases (the most complex 1/3?) will probably remain beyond autonomous robots for another 2-3 decades, and possibly longer.
Second, surgical expertise contains a large amount of tacit knowledge. Skilled surgeons do not necessarily have conscious access to, or the ability to formalize, every judgment and adjustment they make during an operation. In one study I discuss in Chapter 16 of my book (Artificial Intelligence: Utopia, Dystopia, or something in between?), experienced surgeons could articulate in advance only ~30% of the cognitive steps they subsequently performed. The rest emerged only through detailed review, and some of it was not consciously accessible even to the surgeons themselves.
This does not mean that robots cannot learn such skills, but it does mean that knowledge cannot simply be extracted from surgeons, written down, and transferred into an AI system. Progress is much more likely to come through narrow, procedure-specific systems working with human surgeons than through a general-purpose humanoid robot replacing them.
People like Musk repeatedly underestimate the complexity of human judgment and experience, as well as the complexity and unpredictability of the real world. The best application of AI in surgery is going to be as an enhancement of human surgeons' ability. For example, as an overwatch that follows the surgeon's work, calls out issues or anomalies, and reminds the surgeon if a step is being missed or an optional task is recommended. It can also enhance the precision with which the surgeon operates (as is already done with machinery or robotics in many surgery applications).
I agree. AI’s most valuable near-term role will be to enhance the surgeon’s judgment, situational awareness, and precision rather than replace the surgeon.
There may be exceptional situations, however, in which autonomous AI-guided surgery becomes justified: for example, in Antarctica, during spaceflight, eventually on Mars, where no qualified surgeon is available and delay could be dangerous. Even then, I would expect this to be limited to relatively simple, highly standardized, and urgent procedures, ideally with remote human supervision whenever possible.
In ordinary clinical settings, and especially in complex or unpredictable operations, a human surgeon will remain involved for many years.
I’m down for watching a livestream of Elon as the first surgical patient.
I mean… it went so well when he demonstrated the indestructible cybertruck glass.
Elon Musk is just an embarrassing human. I have no idea how much of his wealth is "deserved", I have no idea how shrewd of a business man he is, I have no idea if he's an elite gamer like he claims (seems about as credible as the claim above), but I do know he's a deeply embarrassing person. Guy needs to stop digging.
>Deserved wealth
He's from a very wealthy family. He started off on a low rung in Silicon Valley and, to his credit, has built up from there and learned from early mistakes. He did start off in a very good position (particularly compared to the median South African).
>Elite gamer
Certainly not. There's scores of YouTube videos showing the direct opposite.
For Musk's adherants, Elon's ridiculous predictions are a feature not a bug.
It is far more likely that car salesmen like Elon will be replaced by robots in the next few years than that surgeons will.