pull down to refresh
Thanks! I didn't like the version before this - somehow it was underperforming for what I needed it for (tested it in February for production use). Will report back when I get a moment/case to check it out for something real. If it's really so much better than GPT-OSS-120B, @bitcoinplebdev may have a potential new fav open source model.
sounds too good to be true?
Hmm no but they omit the time he and the people that volunteered to help put into it. It wasn't just $3000.
Here's the original press release: https://news.unsw.edu.au/en/paul-is-using-ai-to-fight-his-dogs-incurable-cancer
His dog is still alive today.
Original press release is from June 2025
Tumor shrunk by half and dog is more active
Exactly, so I don't think at all that it is too good to be true. Just it's not 1 query to chatgpt, 1 query to alpha fold, pay $3k, done. So you have to commit to it and navigate a really complex material, be smart enough to validate what the models say, make choices, understand the subject matter. I feel that there's downplay on these things in popular narratives out there.
But, yes. Much things are in reach now.
I'm still skeptical of mRNA or messenger RNA vaccines
Part of it is my animus toward Moderna
@optimism