I think migrating away from office is not realistic for many companies. They run on Excel, and just breaking one vital table because it uses a macro can cost much more money than the entire office suite for many users…
It is not very good at hard tasks, its ranking is much worse there.
The next question is if LLMs are actually more sexist than the average human working in HR. I am not so sure...
No, for this you need https://github.com/rmcrackan/Libation
Grok-2 is rank 2 on LLM arena, it's basically as good as the best Gemini model. They already caught up. Only the latest ChatGPT model is a tiny bit better.
I once wanted to compile a program that used Xapian on Windows. It was basically impossible for mortals. Imo people should use cross-platform alternatives.
It will likely be amazing, Sam Altman said that the step between 4 and 5 will be like the one between 3.5 and 4. You can of course doubt him, but we'll see... I guess it will be this year, some guy working at OpenAI…
>executing large-scale changes in entire repositories in 3 years You can look at SWE-Agent, it solved 12 percent of the GitHub issues of their test dataset. It probably depends on your definition of large-scale. This…
It is impossible to remove personal data ("any information which are related to an identified or identifiable natural person") from the LLM training data. As far as I understand it ChatGPT and all other similar systems…
You can use DBeaver to get a good diagram for almost all RDBMs.
Everybody who has released a decent model has used a shadow library dump for training. Are they going to admit that they illegally downloaded more than 20 million books?
Scholz does not stall support for Ukraine, Germany has given about 10 times as much military aid as France to Ukraine: https://www.politico.eu/article/military-aid-ukraine-france-... Also compared to the US, Germany has…
Do you mean direct air capture (DAC) or carbon capture and storage from burning fossil fuels?
He's talking about the new C# extension, Pyright and Pylance are free for everyone.
The head of the Eclipse foundation said that they are aware, but that they thought that it would be too expensive for small businesses to get the verification. So the open source devs should do it for free. Related…
When you look at open source AI, for example, EU laws are a huge problem: Not only the act you mentioned, but also GDPR, where it is not clear so far how the "right to be informed" or the "right to deletion" can/must…
He apparently got divorced and moved out of his home country. And here he has no social circle. There are of course hopeless cases, but he is not mentally ill, only alcoholic. It is possible that he will never overcome…
There are very few real-world-tasks where GPT-4 is worse at reasoning than the average person, right now only long-term memory, latency, TTS and STT makes virtual people suck. Once that is solved we should of course…
Decent local LLMs such as Vicuna-7B are already able to run on phones with 8 GB RAM: https://mlc.ai/mlc-llm/ Right now it's a prototype, but in the future local LLMs will be everywhere.
Currently many for-profit companies try to do the same, but additionally emotionally manipulate the user into giving them as much money as possible. (To not even mention automated catfishing...) This project is much…
It is a community newspaper that is (apparently) independent of the WMF, and they don't have quarrels about hosting WMF-critical articles. Anti-WMF sentiment is surprisingly common on some Wikimedia projects.
I think this applies to people with a university degree, it is rare for them to speak bad or no English. (Many speak it absurdly well.) But an average young nurse will probably only speak bad English and can maybe say a…
Similar to ChatGPT it also fails the "What is heavier, one pound of feathers or two pounds of lead?" test. So far only GPT-4 passes that one.
I could replicate the refusal, but all it took to get the model to reply properly was replacing "evil hacker" with "hacker". Paternalism is not that big of a deal right now, especially when compared to ChatGPT/Bing…
I think it is a great idea. The problem is that institutions/decision makers have very little knowledge about open source software so it is really hard to convince them to do this. I can only speak about Germany, but I…
I think migrating away from office is not realistic for many companies. They run on Excel, and just breaking one vital table because it uses a macro can cost much more money than the entire office suite for many users…
It is not very good at hard tasks, its ranking is much worse there.
The next question is if LLMs are actually more sexist than the average human working in HR. I am not so sure...
No, for this you need https://github.com/rmcrackan/Libation
Grok-2 is rank 2 on LLM arena, it's basically as good as the best Gemini model. They already caught up. Only the latest ChatGPT model is a tiny bit better.
I once wanted to compile a program that used Xapian on Windows. It was basically impossible for mortals. Imo people should use cross-platform alternatives.
It will likely be amazing, Sam Altman said that the step between 4 and 5 will be like the one between 3.5 and 4. You can of course doubt him, but we'll see... I guess it will be this year, some guy working at OpenAI…
>executing large-scale changes in entire repositories in 3 years You can look at SWE-Agent, it solved 12 percent of the GitHub issues of their test dataset. It probably depends on your definition of large-scale. This…
It is impossible to remove personal data ("any information which are related to an identified or identifiable natural person") from the LLM training data. As far as I understand it ChatGPT and all other similar systems…
You can use DBeaver to get a good diagram for almost all RDBMs.
Everybody who has released a decent model has used a shadow library dump for training. Are they going to admit that they illegally downloaded more than 20 million books?
Scholz does not stall support for Ukraine, Germany has given about 10 times as much military aid as France to Ukraine: https://www.politico.eu/article/military-aid-ukraine-france-... Also compared to the US, Germany has…
Do you mean direct air capture (DAC) or carbon capture and storage from burning fossil fuels?
He's talking about the new C# extension, Pyright and Pylance are free for everyone.
The head of the Eclipse foundation said that they are aware, but that they thought that it would be too expensive for small businesses to get the verification. So the open source devs should do it for free. Related…
When you look at open source AI, for example, EU laws are a huge problem: Not only the act you mentioned, but also GDPR, where it is not clear so far how the "right to be informed" or the "right to deletion" can/must…
He apparently got divorced and moved out of his home country. And here he has no social circle. There are of course hopeless cases, but he is not mentally ill, only alcoholic. It is possible that he will never overcome…
There are very few real-world-tasks where GPT-4 is worse at reasoning than the average person, right now only long-term memory, latency, TTS and STT makes virtual people suck. Once that is solved we should of course…
Decent local LLMs such as Vicuna-7B are already able to run on phones with 8 GB RAM: https://mlc.ai/mlc-llm/ Right now it's a prototype, but in the future local LLMs will be everywhere.
Currently many for-profit companies try to do the same, but additionally emotionally manipulate the user into giving them as much money as possible. (To not even mention automated catfishing...) This project is much…
It is a community newspaper that is (apparently) independent of the WMF, and they don't have quarrels about hosting WMF-critical articles. Anti-WMF sentiment is surprisingly common on some Wikimedia projects.
I think this applies to people with a university degree, it is rare for them to speak bad or no English. (Many speak it absurdly well.) But an average young nurse will probably only speak bad English and can maybe say a…
Similar to ChatGPT it also fails the "What is heavier, one pound of feathers or two pounds of lead?" test. So far only GPT-4 passes that one.
I could replicate the refusal, but all it took to get the model to reply properly was replacing "evil hacker" with "hacker". Paternalism is not that big of a deal right now, especially when compared to ChatGPT/Bing…
I think it is a great idea. The problem is that institutions/decision makers have very little knowledge about open source software so it is really hard to convince them to do this. I can only speak about Germany, but I…