@iriepixel the answer is quite simple. Free speech and reach. Mastodon is fragmented and each instance has its own sets of rules, admins block entire instances easily, and there is no way to appeal that.
Of course it's every admin's right to manage their instance however they want. But the result isn't desirable for many people.
There are also other smaller issues that can ruin UX.
I hate politics and I generally try to avoid positing about it for my own peace of mind. But it's nauseating to see how some people in the world refer to him as "president". He was as "president" and came to power as "democratically" as all those who sent their condolences including but not limited to:
@Mehrad
He was the president, no matter how much we hate that. Is the process truly democratic? Absolutely not. But that's not relevant. Name of that political position is the presidency.
US government also offered condolences, as well as Japan and Jordan and some other countries. You can't cherry pick ones that fit your argument.
P.S.: I am NOT defending him, or regime in any shape or form.
Generative AI epitomizes the worse tendencies of our current society. It’s simply not worth the cost and only benefits a minority when the majority is paying the cost in terms of environmental, societal and mental wellness impacts.
The efforts spent on generative AI would probably be better served on solving actual problems instead of faking human competency but it’d be too easy if that was something techbros could fathom.
@metacosm unfortunately you and many other people are only focusing on proprietary services and completely ignore what's happening in open source generative AI community.
I only wonder if it's intentional or comes from lack of knowledge.
The hype and utility about LLMs are overstated and will cause problems due to leadership teams in organisations (and Governments) buying into the hype.
However, they do have value as personal assistants, research assistants, and sounding boards as long as you treat all LLM output critically, especially on topics where you are not an expert.
I’m using Claude 3 Opus as a research assistant. It’s read more of the world’s info than I ever will. I am also trying out ChatGPT-4o.
@ianRobinson Among the commercial ones that are available online, Claude is definitely the best one.
Though I personally run LLMs on my own PC 90% of the time.
Hands down, the best font I've found to have high readability, low eye strain, and at the same time beauty is Fira (has both Sans, Mono) and it developed by Mozilla:
Also in some IDEs use Fira Code which is separately developed but is based on/inspired by Fira, but it also supports tons of handy ligatures that not only makes some programming or markup languages more easy to read, but it is also very legible and beautiful:
@Mehrad in case you're curious how they look with Iosevka.
Though Iosevka has many different variants and is highly customizable. (I use my own custom Iosevka)
All it would take for AI to completely collapse is a ruling in the US saying these companies have to licence the content they used to train these tools.
They simply would never reach a sustainable business model if they had to fairly compensate all the people who wrote, drew, edited, sang or just created the content they use.
Simply being forced to respect attribution and licenses would kill them. Will that ruling ever happen? Maybe not. Should it? I think so.
@thelinuxEXP honestly, I don't think that's necessary. Training a LLM isn't the same as using copyright materials. That's like saying if I copy paste your this post into a text file on my computer requires me to pay you for it!
Instead, I'd argue to give incentives to companies to release their LLMs publicly, Like Meta and Mistral do.
Unless you are truly looking for killing generative AI, in which case, we can't have any discussion. But I can say throughout history, every new tech had faced people who thought it was their duty to destroy that technology no matter the cost.
@ErikUden That's not what fedidb.org says. Probably this bot also counts all the accounts and servers that have been long deleted as well. Not number of current accounts.
Also, what matters the most is number of active users, which is just about 1m.
I downloaded and tried Dolphin 2.7 Mixtral 8x7b today. So far, it's performing well. It's not too slow, averaging around 7.5 tokens per second. Tomorrow, I plan to test its performance in different topics. Perhaps I should also compare it to the other models I've tried. #LLM#LocalLLaMA#AI#GenerativeAI
@nosherwan
There's a lot of models you can use. You should just try some and see which one fits your needs best.
Beside this Dolphin Mistral, I've also used Wizardlm 13b, Mistral 7b and Nous Hermes 13b. They've all been decent. And you should be able to run most 7b or 13b models without issue.