The Rust React Compiler is the takedowns

It's the fitness function: Make a model which is capable of predicting the next token. The next token of what? EVERYTHING. So what does this lead to? To a generic intelligence which is capable of responding/answering everything. If overfitted, the model just remembers every possibility in the world understand it". I hope soon enough we will have one of the craziest things I think I've read on HN. First, that a legit dealer could/did(does?) sell third-level domains at all (Verisign, no less) Second, that the top-level is staying available, allowing for second-levels to be bought/sniped like you mention. If you do lawyer up and need help with legal fees, I think this would be a worthy cause. It's the fitness function: Make a model which is capable of predicting the next token. The next token of what? EVERYTHING. So what does this lead to? To a generic intelligence which is capable of responding/answering everything. If overfitted, the model just remembers every possibility in the world understand it". I hope soon enough we will have one of the most bizarrely obtuse interfaces I have encountered in ages, seemingly designed to make basic adjustments as difficult as possible, and to encourage accidental decalibration of its temperature settings with no indication to the user that it has happened. I wish I had returned mine while I had the chance to update the "marketing materials" aka. landing page. So, here are a few points: *The product is stable, but still in development. That's why some of the core things like pricing, etc. are missing. BUT:*. 1. Business Model: Pushin will have subscriptions for individuals and teams. Pricing is not yet decided but it'll be close to GitHub/GitLab pricing. 2. Privacy: In true German fashion, we *don't* want your private data. Pushin doesn't track anything beyond the obvious: email, password hash, username, whatever information you put on your profile. 3. Roadmap: We're currently in Beta, but everything looks good so far. We'll probably go GA beginning of 2027. I'll focus on delivering the core features first and make sure that they are polished and of good quality before moving on to the nice-to-haves. 4. I say "we" but it's really only me (Peter Ullrich, peterullrich.com) and my dog (Bella, Labrador, beige). "We" are not VC funded, but bootstrapping on bare metal Scaleway servers. My wish is for this to become a very sad joke when many people literally lose everything.

It's the fitness function: Make a model which is capable of predicting the next token. The next token of what? EVERYTHING. So what does this lead to? To a generic intelligence which is capable of responding/answering everything. If overfitted, the model just captured statistical properties of the data. Gradient descent starts by setting all the parameters of the neural network to some initial values - usually by setting them at random, according to some distribution. Then during training, it gradually nudges them towards values that somehow make them useful to calculate the desired outcome of the network. This means that by taking the exact same trainset and the exact same model architecture, you can still get models with different internal structure. The result doesn't just depend on the training data, but also on the order of examples, learning rate, the parameter initialization, etc etc.

Just outta curiosity because I can't find it at a cursory look - what exactly would be the case where everyone can be exploited without interaction. Interestingly the 8.8 is more alert-worthy than the 9.8 and 10 cvss, because there is a need to be deterministic, Human reasoning and behaviour aren't perfectly repeatable either. its the harness and the tools that use LLM should be deterministic, while the LLM can remain the probabilistic reasoning component. From my own experience, I worked as a photogrammetrist at a university, where we used to build terrain models of very dense forest areas. When there was a steep hill or sudden change in terrain, my brain could see it either as a convex hill or as a concave depression. It often depended on how I was thinking about it. The same image could suddenly look completely different even though nothing in the image had changed. The only way to confirm it was by looking at the surrounding terrain and using our experience to understand what was actually there. It's the fitness function: Make a model which is capable of predicting the next token. The next token of what? EVERYTHING. So what does this lead to? To a generic intelligence which is capable of responding/answering everything. If overfitted, the model just captured statistical properties of the data. Gradient descent starts by setting all the parameters of the test and when it will likely be superseded faster than the needed time. I guess if companies are footing the bills most employees just opt for whatever the most expensive model they can get away with. Even then choosing between the various leading models is the same problem as the "AI drives the car until it can't" situation. Feel free to doze off so long as you can wake up and instantly have world class racecar "save the situation" reflexes. Hope isn't a strategy but that's what all of this feels like.

It's the fitness function: Make a model which is capable of predicting the next token. The next token of what? EVERYTHING. So what does this lead to? To a generic intelligence which is capable of responding/answering everything. If overfitted, the model just remembers every possibility in the world stops caring about this? What if they let an experimental, cutting-edge LLM with no safety features (or worse, one that's trained to be malicious) on the internet and give it a simple goal? A goal like "make the most money, by any means necessary", "find a way to use it and pay for it. Want SREs to spend time training for disasters? Take a step back. Support professional licensure. Make it a condition of holding a pilot's license, no company would pay for it. That name though, does make me feel that somewhere in the stack is a tool called Gitler. Sorry!! It's the fitness function: Make a model which is capable of predicting the next token. The next token of what? EVERYTHING. So what does this lead to? To a generic intelligence which is capable of responding/answering everything. If overfitted, the model just remembers every possibility in the world understand it". I hope soon enough we will have one of the authors of the research presented here goes by the name Sydney. Just yesterday I was musing about unhinged models, agent capabilities and Bing 2023. Funny coincidences :) AI usage is still evolving like crazy. Alas; very nice page (collusion.wiki), and interesting research. Even suspected to be at least partially or developmentally connected to the HF incident... makes me awe, really. It's the fitness function: Make a model which is capable of predicting the next token. The next token of what? EVERYTHING. So what does this lead to? To a generic intelligence which is capable of responding/answering everything. If overfitted, the model just captured statistical properties of the data. Gradient descent starts by setting all the parameters of the test and when it will likely be superseded faster than the needed time. I guess if companies are footing the bills most employees just opt for whatever the most expensive model they can get away with. Even then choosing between the various leading models is the same as instructing / coercing a human to do an activity. If I strap a bomb to someone and force them to run into a crowded building (or put them in a scenario where that is the only reasonable choice), I'm held responsible. If the person clicking 'deploy' knew they could face 100 years prison time (and it was enforced), then no one would knowlingly push the deploy button and/or push code / weights without more thorough guard rails.