If the report is predicting 100 years in the future, are we really "to the brink" now? Not to mention the decay time of these LEO is in months, hardly a kessler syndrome scenario.
The report says we will be in trouble in 10-20 years, not 100. Look under "collisions causing collisions". Debris from previous collisions increases the risk of future collisions.
It probably doesn't matter. Global warming has proven that we are incapable of doing the necessary things to slow or stop the human-caused decaying of our planet. I'm starting to think we deserve what we're gonna get.
> If we are on the brink of Kessler Syndrome in LEO I would expect market pricing to reflect that.
Wouldn't the market pricing only reflect this after that fact? Up until it happens and no one knows any better, business will continue as usual because otherwise numbers don't go up. Like any past financial crisis, things look fine until they don't.
If you think anything Elon doing is groundbreaking, you have no idea how the world works. Recent Space X ipo showed that the launches aren't cheaper, they are just heavily subsidized. Tesla was a piece of crap until they got their model 3, the only reason Tesla succeeded with their S model is because Elon was the edgy hype dude who managed to generate enough hype to carry them through the bullshit with the car. Self driving was supposed to be solved last year, and tiny companies like Comma AI manage to build self driving systems that are in someways better than Teslas.
I bet you think Steve Jobs was a visionary as well lol.
>If not for Tesla there won't be EVs you see today.
Most manufacturers already were working on hybrids, which to this day are still suprerior to EVs. Chevy Volt, outside of being Chevy, was still one of the best cars ever made for utilitarian purpose. Nobody wanted to foot the bill to do electric conversions until this was necessary.
Tesla only opened up a market segment for high end electric cars, which I guess is cool, but far from revolutionary. The model 3 was a big success only because again, it was subsidized. Meanwhile BYD actually makes cheap affordable electric cars, and we both know why they are not sold in US.
>If not for SpaceX there wouldn't be gigabit internet connectivity in the middle of the ocean.
Plenty of companies were doing geostationary orbits with satellite connectivity. SES for one.
Any more Elon slop? You realize you are defending a dude that is literally a Nazi, right?
My response was to your statement “Give me one actual reason why a smart person would take an underpaid overworked position at any of his companies.”
I hated seeing him do the “Roman” salute and will not try to defend it. I said nothing in favor of his politics or personality, only that his companies are ambitious. Seems like you may benefit from a few deep breaths before responding with attacks.
Also, the alternative technologies you propose are pretty weak comparisons as they are of a different classes.
Your response was along the lines of "despite Musk's character, his companies do groundbreaking stuff, so in return for being overworked and underpaid, they get to work on things that no one else is working on"
My point is that none of the companies do groundbreaking stuff, and there are other options for people wanting to work at those.
I guess we’ve found the point of disagreement then. If you are disregarding the scale, then you might be able to make a case, but even then, I’m fairly certain that you would still be wrong.
Currently, nothing that any of Elons companies are doing is scaling. I don't care if someone back in 2018 wanted to work at Tesla, because of how well Model 3 was doing. Today, the colossal turd that is the Cybertruck is their flagship product, Model 2 is nowhere to be found, Cybertaxis are dangerous, and Roadster is never happening.
For SpaceX, their launches are just as expensive (if not more expensive) compared to traditional methods, they are just subsidizing it, while burning money on Starship that will never reach Mars.
So again, no smart person will ever work of any of Elons companies today based on your criteria.
> More than 4,400 current and former SpaceX employees are likely to become millionaires in the I.P.O., according to an analysis by Hill.com, a San Francisco-based investment platform. Of those, about 400 are expected to earn $100 million or more.[1]
Not to mention the whole launching reusable rockets thing which is pretty cool too.
I mean maybe the super smart person will work at GOOG/META/Netflix to do some absolutely ground breaking work stuffing ads and slop pixels into as many ocular nerves as possible.
However, I could see a smart person taking an "underpaid" (SpaceX SWE: ~$200-400k pre-IPO) to do the following:
Neuralink: to see the impact of increased independence and autonomy of a paraplegic one day after the operation
SpaceX: to quite literally approach the final frontier. Currently launches 80-90% of all orbital mass. Starlink is saving lives constantly.
Tesla: to kickstart the EV revolution and reduce fossil fuel dependence
Boring Company: to radically decrease tunneling costs applicable to all sorts of critical urban problems from transportation to utilities etc
Having listened to him give a couple of talks I am always struck by how much he talks and writes like Claude code.
I don’t think he farms out all of his writing and talking to LLMs. I don’t think Claude code was trained to emulate him or anything.
I think his “voice” has been filed LLM smooth by years of agent based interactions. He claims Anthropic engineers use an average of 500+ agents a day. They are human interfaces to token generators more than human to human communication. They are picking up the tendencies of their most frequent communication partner.
Unfortunately, I see Claude being wrong often enough that I view it as a faulty narrator. Often helpful, sometimes totally full of it.
And now I subconsciously apply this filter to anything that sounds like Claude.
Makes me nervous that my voice may be becoming that of a faulty narrators.
>"They are human interfaces to token generators more than human to human communication. They are picking up the tendencies of their most frequent communication partner."
I see this happening at work, too. All communication and collaboration between engineers has broken down. Juniors don't have questions anymore. Seniors don't discuss architecture anymore. Everyone is just off in their own world of feature work with the only interaction coming at merge time. Really depressing what has happened to our trade.
Higher level thinking has fewer constraints on being wrong than lower level subconscious thinking. Low level stuff tends to have some repetitive and evolutionary basis. Higher level thought tends to be more one off, less evidence based, and more exploratory until we get to the point of being an expert around particular concepts.
A huge amount of human thinking is being a parrot and repeating statements without a deeper understanding. For example when I say 1+1=2 I'm not thinking about it. If I say 638483+949948= calculation is necessary.
So in this, your voice has always been one of a somewhat faulty narrator, LLMs just may be further degrading the quality of the narration.
Lets say you have a colleague named Bob. Bob is a pretty good engineer, but he is as confident when he doesn’t know as when he knows.
You can never tell his degree of confidence in his answer. For low-level things he is pretty good but for critical high-level things it is more ambiguous.
You work around this by checking Bob’s work. Having formal verification steps to check what he says.
Now let’s say you meet Bob’s twin brother Bill. Who talks and sounds just like Bob. He makes fairly grand high-level statements that you can neither test nor verify.
Bob also knows how to quickly check things with a little bash scripting and tests. Can quickly switch branches to confirm regressions and consult documentation if something requires more information. His twin can also quickly scan for all the common security vulnerabilities in any code, and will do it without complaining as many times as you want him to.
I’ve come to trust Bob with code more than any human.
> I think his “voice” has been filed LLM smooth by years of agent based interactions.
I've wondered this for a long time now: how does one differentiate between your voice getting "filed smooth to match LLMs" or your brain getting "filed smooth to match LLMs"?
After all, if we accept (as the token providers want us to) that evidence of language use as we see in LLMs is evidence of intelligence, then surely the reverse is true - evidence of machine-only language use is evidence of machine-only intelligence.
IOW, it's the LLM talking through you, not you talking through it.
The problem here is that the latest Claude models have gone off the deep end in their communication, and there seems to be neither awareness nor willingness to fix this. "Having your writing and talking turn into Claude is the risk nothing names" or what vapid strange meaningless thing would Opus say.
There might be some confusion on your part on just how much AI compute has been installed in the past few years in relation to the total amount of CPU compute and the speed at which it was installed.
If the CPUs life was an hour, the GPU AI accelerator would have been around 5 minutes and is around 5x larger in total compute.
It should not be underestimated in any way the scale of what is occurring.
No, I think the confusion on your part is around what it sounds like when you say "in relation to human timelines". Humans have been here for around a million years. We've been living in cities for several thousand. It sounds like what you're actually comparing to is AI computing to pre-AI computing, which is reasonable, but I don't really see how that is accurately described by "in relation to human timelines".
That isn't humanly possible, you will blow out your neurotransmitters and not be able to think. Only being semi serious, but we can't make that many decisions a day.
That isn't typically how abstraction is tallied. The ceo doesn't say they use 15k employees, or the programmer uses 30B transistors. Or the director uses 120 workers. You enumerate what you actually interact with.
What does that mean though? Does it mean there are more than 500 different "agent" systems within Anthropic and employees interact with most of those in a daily basis?
Any court case involving an LLM you can point to where LLM output violated copyright AND that was held up in court?
General thought here is the entire economy is all in on a bet that that would never be enforced and I am willing to bet we would rather change the law than destroy the economy…
We have actually little idea how human brains work.
Every age thinks they know how they work, and then every subsequent age laughs at the previous age’s rudimentary understanding.
We gave a Nobel prize to the psychiatrist for inventing a method to remove the frontal lobes of a brain through the nose in 1949. Lobotomies were performed through the 70s.
We are likely doing similar if not subtler but worse things today (just one more pill bro). We still have no idea what we are doing when it comes to the brain.
From a very naïve perspective, why isn’t air traffic control a solved problem?
Would love to be educated here, but it seems like there are just not that many moving pieces per airport (<10,000) and all of them are tagged with transponders, etc. And a human in the loop with respect to coordination might be the biggest source of risk.
This all seems like queuing theory and collision avoidance, which are very well done by computers.
Worked in this area for a couple of years. We have automated collision avoidance systems and have for years. But things can go off nominal fast and you don't want passenger aircraft making drastic maneuvers that fast. Because conditions can change so fast, queueing theory isn't that useful either. It's useful for more strategic as opposed to tactical air traffic control, what we call air traffic flow management. There's been a ton of research in this area for decades, for example at the National Center of Excellence in Aviation Operations Research (NEXTOR) and at NASA's Aviation Systems division. I could say a lot more offline about why, in my humble opinion, progress has been unbelievably slow, but the short answer is that it's government plus modern day academia operating in an environment where funding is low and risk tolerance is lower.
I'd suggest looking at Admiral Cloudberg's recent analysis of the Jan. 2025 airliner/military helicopter collision in Washington DC. It's a study in the difficulties of air traffic control in the 21st century and the reality of transponder usage, complicated by an overwhelmed airport and legislative indifference. Transponders aren't universally used, technology and people aren't perfect.
This one is long; I suggest reading the first chapter, then skipping to chapter 7 for the important parts of the story.
Weather conditions, ground/runway conditions, obstructions, mechanical issues. There are tons of edge cases with huge consequences for error. Same reason self driving is not a solved problem, but with dramatically less tolerance for error.
But it feels like planes have a huge advantage in that they all have trackers constantly broadcasting their location to one another. Almost like they are working as a team compared to FSD systems which are limited to a small range of self collected info from vision/lidar.
Not that it can ever be fully automated but it seems like much easier problem to solve.
Doesn’t meaningfully change the answer. There are still tons of edge cases with catastrophic failure modes. The transponder won’t tell you about the alligator on the runway, a vehicle where it shouldn’t be, unexpected braking action, a mechanical problem, or a plane not doing what it was cleared to do. Tracking the planes is the easy part.
The reasons we haven't automated air traffic control are similar to the reasons why we still have pilots in the cockpit even though we technically don't need to some of the time.
The thing is... not everything you need to track, not even all airplanes, will have transponders.
Sometimes the transponders will be faulty, sometimes you have to direct a broken aircraft safely to the ground with no transponder available.
And sometimes you are doing ATC for a zone that does not, in fact, require transponders (those are required only for the strictest controlled zones, not in general)
My impression is that it's at least partially down to how brittle a fully automated system could be. You want as much of a "manual" backup as possible, and that only be achieved by doing things manually all the time to keep up practice. With the current system, even with enormous degradation (massive power outages, cyberattack, software flaw, etc.) it's possible to get planes out of the sky without much risk. If a fully automated system fails, it's an enormous disaster.
The other problem is that the "worse is better" arrangement where ~everything is done via voice over radio makes automation difficult. Voice is simple, universal, and enormously flexible, but hard for computers to handle systematically. Some two-way protocol and associated pilot UI would be useful... but then you have to decide on that protocol and UI, every manufacturer has to implement it, every airline has to buy it, and every pilot has to be trained. Unless it's very, very clear what the optimal solution is and that the benefit outweighs the cost, that sort of coordination is difficult.
(And remember, the ultimate result of such a change would be a rigid system that is difficult to change, more fragile, and does not save much money relative to the total cost of the airline industry.)
It's not only a dangerous assumption, it's simply incorrect. Transponders on aircraft are not required in most of the US, including around Class D towered airports unless operating within a designated Mode C veil or at/above 10,000 feet MSL.
Consider the ratio of air traffic controllers to the number of people moving through the airport. It's not a huge cost. There's often far more staff moving bags.
It was a problem at very rural airports, and now some people man those remotely.
Apple leaks like a sieve. The extreme secrecy culture is a pointless drag on productivity, maintained long past its usefulness for the sole benefit of the execs practicing their Steve Jobs "one more thing" keynote cargo cult.
I would be honestly intrigued but wary of just the usual Elon bad/good so all things he does bad/good narratives.
If we are on the brink of Kessler Syndrome in LEO I would expect market pricing to reflect that.
reply