7
points
birktj
Joined 286 karma
- birktjThe argument is that scams based on exploiting goodness causes a lot more harm compared to the ones based on exploiting greed. Because it trains people that doing good deeds is not worth it (they might be scammed.) And even if the rate of such scams are low, just reading about them makes people afraid of potential consequences of doing good deeds. So I absolutely agree that such scams should have very harsh punishments, because they do not only have immediate consequences, but they degrade trust in our society.
- I strongly disagree and I am having trouble understanding what kind of world you envision, what will it look like?
The problem as I see it is not robots coming for my job and taking away my ability to earn a salary. That can be solved by societal structures like you are saying, even though I am somewhat pessimistic of our ability to do so in our current political climate.
The problem I see is robots coming for my mind and taking away any stakes and my ability to do anything that matters. If the robot is an expert in all fields why would you bother to learn anything? The fact that it takes time and energy to learn new skills and knowledge is what makes the world interesting. And this is exactly what happened before when machines took over a lot of human labour, luckily there were still plenty of things they couldn't do and thus ways to keep the world interesting. But if the machines start to think for us, what then is left for us to do?
- Absolutely, I am sceptical of AI omin many ways, but primarily it is about the AI companies and my lack of trust in them. I find it unfortunate that all of the clearly brilliant engineers working at these companies are to preoccupied with always chasing newer and better model trying to reach the dream of AGI do not stop and ask themselves: who are they working for? What happens if they eventually manage to create a model that can replace most or even all of human computer work?
Why whould anyone think that these companies will contribute to the good of humanity when they are even bigger and more powerful, when they seem to care so little now?
- To a certain extent I agree. But as far as I know I cannot go to chatgpt.com and paste the newest ICPC problems and get full solutions. And there is no information about what they do differently. For a competition like the ICPC, which is academic in its nature, I think it is very unfortunate to setup a seperate AI track like this without publishing clear public information about what that actually entails. And have clear requirements for these AI companies to publish their methology. I know it is a nice source of sponsorships for them, but the ICPC should afford to stand up a bit for academic integrity.
Without any of this I can't even know for sure if there was any human intervention. I don't really think so, but as I mentioned the financial pressure to perform well is extreme so I can totally see that happening. Maybe ICPC did have some oversight, but please write a bit about it then.
If you assume no human intervention then all of this is of course irrelevant if you only care about the capabilities that exist. But still the implications of a general model performing at this level vs something more like a chess model trained specifically on competitive programming are of course different, even if the gap may close in the future. And how much compute/power was used, are we talking hundreds of kWhs? And does that just means larger models than normally or intelligent bruteforcing through a huge solutionspace? If so, then it is not clear how much they will be able to scale down the compute usage while keeping the performance at the same level
- They apparently managed gold in the IOI as well. A result that was extremely surprising for me and causes me to rethink a lot of assumptions I have about current LLMs. Unfortunately there was very little transparency on how they managed those results and the only source was a Twitter post. I want to know if there was any third party oversight, what kind of compute they used, how much power what kind of models and how they were set up? In this case I see that DeepMind at least has a blog post, but as far as I can see it does not answer any of my questions.
I think this is huge news, and I cannot imagine anything other than models with this capability having a massive impact all over the world. It causes me to be more worried than excited, it is very hard to tell what this will lead which is probably what makes it scary for me.
However with so little transparency from these companies and extreme financial pressure to perform well in these contests, I have to be quite sceptical of how truthful these results are. If true I think it is really remarkable, but I really want some more solid proof before I change my worldview.
- Buypass provides ACME certificates as well [1]. The usage limits are not quite as generous as LE, but they work pretty well in my experience.
[1] https://www.buypass.com/products/tls-ssl-certificates/read-m...
- Thanks for the link! I have been thinking a lot about topics related to this lately and this looks like an interesting read. Skimming through it I might not agree with all the ideas of the author I very much welcome new ideas about how to organize our society and our legal system. I have especially been thinking about how the legal system is mostly fiction and what makes it work is not so much what the laws say and what the punishments are for breaking them, but rather our shared belief in the system. And I think this is problematic because it takes away our agency to imagine other ways of doing things, we think our current legal system is the only possible system because it has to be. Thoughts of anything else would cause a collapse.
I have also been thinking about how basing morality on following laws is a bad idea. You should not be afraid of breaking laws or rules if they are stupid. All of modern workers rights, women's rights, etc are based on brave people of the past that were not afraid of breaking stupid laws of their time, why should today be any different.
- I was wondering if anyone here has experimented with running a cluster of SBC for LLM inference? Ex. the Radxa ROCK 5C has 32GB of memory and also a NPU and only costs about 300 euros. I'm not super up to date on the architecture on modern LLMs, but as far as I understand you should be able to split the layers between multiple nodes? It is not that much data the needs to be sent between them, right? I guess you won't get quite the same performance as a modern mac or nvidia GPU, but it could be quite acceptable and possibly a cheap way of getting a lot of memory.
On the other hand I am wondering about what is the state of the art in CPU + GPU inference. Prompt processing is both compute and memory constrained, but I think token generation afterwards is mostly memory bound. Are there any tools that support loading a few layers at a time into a GPU for initial prompt processing and then switches to CPU inference for token generation? Last time I experimented it was possible to run some layers on the GPU and some on the CPU, but to me it seems more efficient to run everything on the GPU initially (but a few layers at a time so they fit in VRAM) and then switch to the CPU when doing the memory bound token generation.
- Why not simply watch native Swedish content instead? There should be quite a lot globally available for free from SVT.
- Related to this Applied Science on YouTube has a pretty cool video where he demonstrates how a single laser diode can be used to measure miniscule vibrations. I wonder if this would be sufficient for a laser microphone?
- GNU/Systemd is pretty hilarious
- Not necessarily, this assumes that it is possible to perfectly simulate physics on computers. It is not obvious that this is true. For one it assumes that physical interactions happen in discrete time steps (or at least be equivalent with a process that happens in discrete time steps). It also assumes that it is possible to perfectly scan the all the properties of some piece of matter (which we know is not possible)
- I'm in Norway so I don't really know how it looks like from abroad, but I am pretty sure a lot of NRK content is available globally. Under the Vikingane show you listed it clearly says "available in Norway". However the first show on the homepage "Nytt på nytt-quiz" [1], a type of game-show, says that it is available globally. I am not sure there is a clear overview over what is available globally though like SVT has.
I think often fiction-type shows like Vikingane has complicated licensing as it is normally produced by some sort of third-party, but other programs like news and the mintioned gameshow is produced in-house and NRK has all the rights. There is a very nice page with an archive of old programs [2], I think most of this should be globally available.
I of course agree in general that geo-restrictions and localized distribution rights are stupid concepts.
[1]: https://tv.nrk.no/serie/nytt-paa-nytt-quiz [2]: https://tv.nrk.no/programmer/nrk-arkivet
- I see most replies here ar mentioning the the build integration is what is mainly missing in the public tools. I wonder if nix and nixpkgs could be used here? Nix is a language agnostic build-system and with nixpkgs it has a build instructions for a massive amount of packages. Artifacts for all packages are also available via hydra.
Nix should also have enough context so that for any project it can get the source code of all dependencies and (optionally) all build-time dependencies.
- 1 point
- Yesterday I was configuring a usb to serial converter to communicate with a CNC controller (over RS232). However I was unable to transmit any data and so I brought the oscilloscope to see what was happening. I disconnected my converter and connected the oscilloscope and all the data was there and my oscilloscope had no problems decoding it with the same parameters i had used with my usb converter. Confused, I disconnected my scope and connected my converter back up again, but sill nothing. After a few times back an fourth I ended up connecting my scope up while the converter was still connected and crucially I also had a cat /dev/ttyUSB0 running and as expected the data was right there at the scope, but to my big surprise it was also there in my terminal! I then tried to disconnect my scope resulting in no more data being sent to my laptop.
The only explanation I could think of was a faulty ground connection. The scope was grounded through the wall socket same as the CNC and thus had the same ground potential. My laptop on the other hand was running on battery and only shared ground with the CNC through RS232 pin 5 (signal ground). However it seems this pin was not correctly connected on the CNC side and thus I was unable to transmit any data. Experimenting a bit I tried to connect ground to the connector shield instead and everything worked perfectly.
- It seems like they are using [1] Wish [2], a library for "SSH apps"
[1]: https://git.sr.ht/~erock/prose.sh/tree/main/item/cmd/ssh/mai... [2]: https://github.com/charmbracelet/wish
- That looks pretty impressive, but I am quite skeptical of the fact that this craft uses ion propulsion for any significant proton of its thrust. A reddit user [1] suggests that it uses ducted fans which I am way more inclined to believe. The very high pitched noise is typical of a high velocity propeller blowing high velocity air. Which seems wildly incompatible with those large grids where one would expect relatively low velocity air with very little noise. There are also four very suspicious tubes running along the body of the craft.
[1] https://www.reddit.com/r/Futurology/comments/o6h1bc/comment/...
- I would recommend people who are looking for not only printers, but most kinds of equipment (except for cases where new stuff is unambiguously better than old stuff, but this is rare) to buy used professional/industrial hardware instead of new consumer equipment. And if this is difficult try to look for old consumer hardware still in good shape. These are often of much better quality and also it is much less wasteful to something used than new.
Professional equipment is usually in another league than consumer stuff when it comes to quality and reliability and also usually has very low resale value so you can get it for quite cheap. Personally I just picked up an old Sharp color laser printer for about 90$. This is a great machine that prints 2 sided A3 (very hard to find for laser printers) and with toners that should last about 10000 pages which is way better than anything you can buy as a consumer. The only backside is that it is quite big and heavy.
Old consumer equipment that is in good shape is often a good deal with the main reasoning being that something that has lasted for a long time has a higher probability of continuing to last a long time compared to something new where it is hard to say how long it will last.