In order to kill all cancer cells in the body, it probably needs to be delivered to every single cell in the organism, and scan the nucleus of that cell. Viruses usually don't infect every single cell, just a small percentage.
So one needs to figure out a delivery method that is efficient enough, and that doesn't elicit an immune response. But I guess one can analyze the cancer in the lab and figure out which receptors it expresses, and then bind to those? We could have a toolkit of different delivery methods, tailored for each patient's cancer.
A new generation of AI companies is out there to take over blue collar jobs as well. Check recent YC batches.
Software engineering was a nice target because inputs and outputs are just data and you don't need to figure out robotics. But idk, 3 years ago it seemed illusory (at least for me) that LLMs could take over software engineering, but now here we are. They are still not 100% there yet (software engineers still have jobs), but we are getting ever closer.
Companies are in the process of figuring out robotics, and even if it's not figured out, then we might introduce a gig-ified blue collar economy where an unskilled, underpaid gig worker implements instructions by AI. Plus a lot of blue collar work already today involves robots (cranes, excavators, trucks, etc).
Robotics aren't new. The LLM robotics trend (half of which are complete scams and the other half are vaporware) might be an even stupider bubble than the LLM programming bubble, though it's also a smaller bubble.
At least LLM programming bubble is applying language models to language tasks, even if the results are mixed. The LLM robotics bubble is doing what exactly? They're making videos of remote-controlled skinsuits doing mundane tasks inefficiently in a way that impressed investors. They're trying to exploit the ELIZA effect for physical movement.
I saw one sorting packages to put the barcode label on top. Do you know what's a better way to do that? You put a camera on every side, including underneath, so the barcode can be read from any direction. This scanner can work at line speed instead of being the bottleneck. This isn't new. And you sort packages into different buckets by having pneumatically activated wedges that swing out and push the package onto a different line. The bottle return machine at my nearest supermarket does that, I'm sure wannabe billion dollar VC funded startups can manage it.
These deals are part of how the AI economy operates. Amodei has explained this in his recent Patel podcast.
1. Building datacenters takes time. Months, if not years. They take billions of investment.
2. AI revenue is highly unpredictable. Sure, you can make predictions, but maybe your competitor releases a better model 2 weeks after your release, maybe the new model you built isn't as much better, maybe the chinese models steal your show, etc.
3. AI revenue grows a lot. Anthropic's case is 10x per year.
4. So if you are off by just a year in terms of how much GPU you actually need, then that means a 90% of your compute capacity is wasted, and you go bankrupt.
As a solution, companies buy compute from each other! If one company's model did well, they can buy compute from the company whose model didn't do well (like in the case of grok). It's beneficial for both sides, so positive sum game. So deals like this aren't something bad in itself.
It's nothing new either. In SAAS deals, you often commit to a certain revenue and then pay extra if your revenue exceeds that amount. And power market is cut in two as well: longer term deals plus spot markets. Spot prices are way higher than the longer term deal prices.
Given it's SpaceX of course there is financial engineering involved: the GPUs aren't actually owned by SpaceX but a daughter company, and it's been financed via loans that are backed by pension funds. So it's already the case that pension funds back bear the risks associated with SpaceX's operations.
Right now, the bulk of the AI bubble sits in such debt statements and not in public markets.
> the GPUs aren't actually owned by SpaceX but a daughter company, and it's been financed via loans that are backed by pension funds. So it's already the case that pension funds back bear the risks associated with SpaceX's operations.
I think a more accurate phrasing of the Valor GPU deal would be something like this:
"SpaceX’s AI compute buildout relies in part on off-balance-sheet or lease-style financing vehicles. Valor-owned vehicles purchased Nvidia GPU infrastructure and leased it to xAI/SpaceX subsidiaries, with Apollo providing debt financing and SpaceX or subsidiaries guaranteeing some obligations. That creates indirect exposure for institutional and retirement capital, though not necessarily direct pension-fund ownership of SpaceX operational risk."
This is only a partial solution to a problem though. If x ai fails to build a good model they rent to Google. But that means all these companies are incentivised to build as much compute as possible. If they win their margins look great, if they lose they still make money. BUT at some point aggregate supply will outstrip aggregate demand and the bottom will fall out of the market.
One of the reasons why over a decade ago, I dived deeply into the OSS world instead of mathematics was that it was so much more accessible: there were docs for everything, and I got direct feedback when something worked vs when something didn't work. Most of my questions had answers on stack overflow, and once I joined Rust (which back then in 2015 didn't have a big stackoverflow presence) I had a community who answered them for me (and in maths I didn't have that).
AI makes the math world more accessible than before. If you have a question about a proof in the lecture, you can just ask it. Of course, one can't trust it blindly, but fundamentally it's amazing.
I think that's a good thing, but of course this means that a lot has to change in culture and behaviors, also in the research world.
The software engineering world is more or less in the same situation, it's also changing. But for now I think it still holds true that someone who knows maths plus an LLM is better than someone who doesn't know maths plus LLM. At least in software it does.
Agreed. As someone who was always curious but had difficulties learning math the way it's taught at the university, AI teaching me the way no professor ever could is a blessing. I fail to see the point of the memo besides: we got here first and we decide what math is because we can. I'm really optimistic about AI and the value it brings in education. Gatekeepers will complain, but ultimately, will either adapt or be left behind.
>AI makes the math world more accessible than before. If you have a question about a proof in the lecture, you can just ask it.
I think that is great, really! but does anyone remember asking a TA or teacher or prof or parent and getting told you can work it out for yourself, or maybe just given a hint? What if that is an essential part of learning, having to work through things you don't understand, but that you have the tools, the foundation, to figure out.
A calculator can't teach you math. A forklift can't build your strength. This is really a double edged sword, as far as education or accessibility goes.
You have to constantly ask... what do I lose by not figuring it out myself?
Yeah, among other factors, that "figure it out" mentality put me off in the end. Especially because often you need to show the same mentality unless you want to overkill proofs and spend more time on them than assigned to you. I sometimes miscalibrated and pointed out some details that didn't need pointing out in my proofs while in other proofs, I skipped over too many details for the TA.
Of course I agree that if the student just asks LLM to do their homework, they have not learned anything. But it's sad if one can't ask questions about a proof or such. Having the LLM around to review the homework submission is also useful, to make sure that the arguments are solid.
You will have to learn to voluntarily figure things out for yourself without being pushed towards that. In a sense it's analogous to the presence of cheap calorie dense foods. In order to not be overweight you have to be mindful of and regulate your food intake in various ways.
Alternatively, perhaps universities will provide access to fine tuned models that are mindful of such things.
Horrible to hear this news. Neurological diseases are the worst because we understand so little about them and usually there is no cure, just management.
What have your experiences been with using AI for medical advice? Especially for such rare diseases I suspect that very little shows up in the training data. Personally I'm using AI only for work and only recently started using it for non-work non-coding stuff too.
> What have your experiences been with using AI for medical advice?
I had been trying to use Gemini during my bout of encephalitis before treatment. I wasn't really trying to diagnose myself, but instead, was looking up side effects of the various (psychiatric) medications I was on. At the time, I (but not my wife) had thought all biological causes had been ruled out due to testing from my PCP. To be clear, I wasn't really in my right mind, so whether this was a reasonable belief or not (likely not) isn't something to be assumed. Like, I just thought I had GAD. Or OCD. Or something latent that had just all of a sudden started rearing its ugly head.
I found Gemini's reporting of side effects of medication to not be helpful. Especially because it led me to wonder if some of the things were "in my head" (without a doctor even needing to say it). Anyway, there was never a point at which any AI suggested anti-NMDA receptor encephalitis. That didn't really come up until I got into the hospital and had an abnormal brain MRI.
I've since switched to ChatGPT, which I find to be leagues better than Gemini personally.
This is all really hard to explain, so I apologize if this doesn't make a lot of sense.
It's fine to make mistakes, that's how you learn. The problem here was that they didn't announce to the host that they are doing a test of their in-development equipment.
So the host wasn't able to add the additional risk and hassle to the price, which in this instance would have been a quite legitimate ask as the robot damaged their revenue generating property.
It's very ironic that Airbnb itself has done similar practices in the past where it ignored hospitality regulations to establish their business model, i.e. not asking for permission but for forgiveness.
The Airbnb style response would be to gig-ify this model where you ask an independent contractor to buy the test robot, rent the Airbnb, and test it out instead of you doing it yourself. Then the contractor bears the risk of damages to the property.
> The problem here was that they didn't announce to the host that they are doing a test of their in-development equipment.
I might be okay forgiving skirting the disclosure rules BUT only if they tried to be model tenants and, if there was any damage, took steps to proactively make things right. If you're breaking the rules, even if there was no damage, you should definitely be cleaning up and putting things back in place.
This was my thought. I can understand not wanting to go to the hassle of trying to explain that you're testing an experimental prototype robot to a confused Airbnb owner.
What I find inexcusable is not owning up to the damage and paying to fix it when your prototype goes on a rampage of destruction.
Moving fast and breaking things is fine, as long as you fix the stuff you break...
Even if it is fixable, there are costs involved for the fixing. A broken hotel lamp will sit in a landfill for all eternity.
"Moving fast and breaking things" could be acceptable in cases where there is an ulterior objective whose potential value could be >> these costs, but in general it should be evaluated more carefully.
In a rental unit you should not have things that can’t be replaced. People who rent it will break things, either by accident or purpose (there are always idiots around).
The problem here was that they didn't announce to the host that they are doing a test of their in-development equipment.
I personally think the problem here is that they were delusional enough to think this was the way to 'test' their prototype clean-o-bots. But as you point out (and...sigh...you're spot on on all points), we live in a world where doing things like beta-testing robo-cars in real live traffic is perfectly cromulent as long as you capture market share and outlast the lawsuits and 'disrupt' something.
Well Bezos did actually state that he wants to turn Earth into a natural park.
But yeah, the robot armies don't need grain so why hike up the price of bread? Lack of grain makes those people resentful which means you need to deal with their anger. Sure, it can be dealt with but it's just cheaper to give the humans grain so they are docile. This is basic governance 101 that goes back to the romans (and further).
They also didn't slaughter all horses immediately. You can't eat that much horse meat anyways. It happened piece by piece.
The only good reason for an abrupt mass culling of the 99% (for a coldly calculating rich person with no empathy) would be game theory, i.e. them not being a contender for power any more. If there are no humans, there is nobody who can question the control of the 1%. It would be thus less about economics and more about power.
I am really rooting for the bottom 99%, myself being a part of it, but I really don't know what will happen to us.
If you give it $290 of input tokens for $10 of output tokens, you are doing something wrong. I.e. you paste the whole CI output into the prompt instead of giving it a link to the file, and then the AI greps its way through it (using a fraction of the tokens).
Sometimes AI overdoes things and it re-runs the whole testsuite because the tail command didn't have enough lines, but the other way round messes up the context so much so that in the end all that context is useless.
So one needs to figure out a delivery method that is efficient enough, and that doesn't elicit an immune response. But I guess one can analyze the cancer in the lab and figure out which receptors it expresses, and then bind to those? We could have a toolkit of different delivery methods, tailored for each patient's cancer.
reply