Plutoberth

> will be available for next to nothing forever

Most people using it for work already pay API rates, and even if you claim API rate inference is not profitable for the frontier models the best open models are still very good, and those are surely served profitably.

This argument is no longer true.

I'm building a game that incorporates LLMs as a game mechanic.

I've been using Luna, but I'll probably switch to Haiku.

See it as $60B value total among Moderna and Merck for the success of the trial, not 150% increase.

A successful phase 3 trial means that the drug is actually effective, not just safe. That's what they test in phase 3.

This is obviously impractical. With the volume of bug reports that are generated and such a wide breadth of software no single agency will be able to handle all reports. And honestly? It doesn't matter, even today. CVE should serve as a reasonably deduped identifier of specific vulnerabilities. It was never interesting and practical to care about 100% of vulnerabilities in a specific deployment, and it's not interesting today.

I'm not sure I understand. Wouldn't any prediction result above statistical random (in the image mind reading study) be significant? If the study was performed correctly I don't really need to know much about fMRI to tell whether it's an interesting result or not.

I don't understand. Wouldn't it only be possible to find out by comparing two identical clocks that were at different altitudes for some larger number of ticks, allowing you to then compare the elapsed ticks? How would you conduct such an experiment? My mental model is that I have a black box that outputs an electrical signal every tick, and then maybe we could just figure out which clock ticked first with a simple circuit. But that seems like we would need to sync them, and that it's fundamentally wrong due to the fact that the information of the tick is also subject to the speed of light. I don't know much beyond high school physics, fwiw.

So the user remains at risk in the time between the leak and the time the company discovers it and resets all passwords, which could be months. It might not really be relevant for most sites and for most users, and you might argue that if the hash database is compromised you have other things to worry about, but it's a something to consider.

But it'll be wrong. 355/113 is only useful up to n digits, not infinitely, and GP is saying that n is as big as the number of digits in "355/113".

Not OP, but I generated over 150 images using DALL-E 2. Results in the quality of the images in the gallery are very common. Usually, for prompts as simple as this most of the output images (there are 4) look as good or better.

172.16.0.0/12 is a private subnet. This means that it's addresses are relevant only within a local network, and never over the internet. If you try to send a packet to an address within that subnet, layer 3 devices (i.e. routers) on the internet will drop it.

If it was that way for a few months and then fixed... still pretty shoddy but sure. However, it has been that way for YEARS, and is one of the most common complaints among players. I wonder how much of the remaining loading time could actually be shaved off if someone with the source code took a crack at it.

Right - thanks for the correction.

It GTFO with its remaining fuel and then crash lands. Going back to orbit would require enormous amounts of fuel.

Starlink roughly orbits at 1200km. The most common orbit for internet sats is geostationary orbit, which means that the satellite is stationary with regards to an observer on Earth. It works because a satellite on a circular orbit at that height moves at the same speed as the rotation of Earth.

This way you can have a simple, pointed antenna and not an obscenely expensive phased array like Starlink. That being said, GEO is only 36000km,so I'm not sure where the "60 times" figure comes from...

This board has four 2.5G ports, not four 10G ports, so the 16G bandwidth for four PCie 2.0 lanes is more than enough.

Failed exploits, especially kernel-level ones, will result in higher system instability. I'm aware of at least one company (ZecOps) that specializes in detecting exploitations using crash analysis.

We're talking about different security models. Whereas anti-cheat companies seek to prevent code execution or introspection in a certain context (the game) by a user that already has full privileges over the machine, other mitigations seek to prevent privilege escalation or initial access. For example (simplification), a remote exploit that relies on a specific address to work would now have to find an additional information leak. Similarly, code running in usermode cannot simply* know the addresses of objects in the kernel.

* In practice this isn't really true, and there are many ways to bypass KASLR