comparity.ai offers every (!!!) important LLM out there for free, including Fable 5 and GPT 5.6-Sol. They can offer it for free becuase its a research project of the Max Planck Institute for Intelligent Systems in Germany. And because its hosted in germany, they take your data/privacy very serious because of german law.
Most innovation (end-user facing) right now is coming from privately owned companies, but theres also a lot of fundamental research going on at public institutions like universities etc. Obviously, companies like Anthropic, OpenAI, Google etc have a LOT more money and are trading the smartest minds like football stars on a transfer market. This all seems kinda absurd, but i guess to make the best products/do the most innovation, you need the best minds.
Then on the other hand, tools like Arena AI came from a research project at a university, and tools like comparity .ai already replicated it (but better, if you trust their researchers, also from a public institution).
It seems to me that at some point, people at public institutions try to pivot into the private sector - because theres simply more money, and public institutions are notoriously underfunded?
How will this shape the future of research, will the standard Bachelors - (masters) - PhD - Postdoc etc path die out in favor of more private-company driven education paths (Even if they cant award degrees)?
https://arena.ai/blog/arena-100m-revenue/
The people over there generate quite some revenue. But with 82M+ total votes, and 100M revenue, i wonder how much profit (or in this case loss) they make.
Hey everyone, Im fairly new in this sub. I really enjoy reading (switched from doomscrolling to books, yey); but often, i cannot enjoy stuff im interested in - mostly popular science books outside of my domain (i work in research)- as I cant enjoy consuming new knowledge anymore after a day of work, which includes a lot of reading.
Do you have recommendations for easy reads for such situations? (Outside of fiction/fun reading)
AI is already the most important tool for drug-discovery (thanks to AlphaFold), is gaining traction in material discovery (e.g. cusp.ai) and probably will be a key driver (or at least critical aid) to solve global challenges like climate change etc. People who fear AI are the same that were hostile towards Carl Benz for inventing this "loud, noisy, unpredictable machine straight from hell" (the car. He invented the car). People feared that they would all loose their jobs when the steam engine became common. When Henry Ford streamlined production, people feared that they were no longer of use.
None of this happened. All those events where central drivers of global development and increased the living standards of hundreds of millions of people.
So im genuinely interested, what (if there is anything) is different this time?
I guess im a little late to the party, i recently read the blogpost that Surge AI postet dec. 25, which is basically a sophisticated way of saying "we hate this". So whats the actual point of Arena? Surge now has their own "expert" rankings, where they pay professionals to rank models, but they didnt publish any data on that (lol). Seems like subjective rankings in general arent really useful besides being a marketing tool