Gemini 2.5 Pro is taking off 🚀🚀🚀
The team is sprinting, TPUs are running hot, and we want to get our most intelligent model into more people’s hands asap.
Which is why we decided to roll out Gemini 2.5 Pro (experimental) to all Gemini users, beginning today.
Try it at no cost at https://t.co/lc7BAJqH5u
When you hear about "human-level AI" being just "one LLM training run away", it's as if people building 3D printers (the kind that makes small, flimsy plastic objects) were telling you that they've nearly built an infinite matter replicator, capable of scanning and cloning any object -- be it steel, gold, or diamond, be it minuscule or building-scale. "We're just one iteration away! Just need to increase spatial resolution a bit!"
As long as you have absolutely no idea what "matter" is and how 3D printers work, it might sound sort of plausible. But if you do, it's absolutely clown shoes. The way to navigate such claims is to have a proper, formal definition of intelligence.
A common misconception about overfitting is that you can only overfit to a dataset if you are directly training on it.
In reality, you can overfit to anything, as long as you have a feedback loop (however small) between results on the target dataset and your model. For instance, if you're basing your model design choices on metrics computed on a validation set, you will gradually overfit to that validation set.
LLMs’ sensitivity to design decisions, such as template choice, can degrade performance & prevent their robust application. Introducing Batch Calibration, a simple method that mitigates this effect & improves on existing methods w/ negligible add’l cost → https://t.co/Y7VbhanA3w
1/6 Thrilled to announce that our paper: “Changes in Research Ethics, Openness, and Transparency in Empirical Studies between CHI 2017 and CHI 2022” has won the best paper award at ACM #CHI2023 !
The recent invasion on Ukraine, our friends and neighbours, left us shocked and outraged.
In solidarity with all victims of this act of aggression, the CD PROJEKT Group has decided to support humanitarian aid efforts by donating 1 million PLN to the @PAH_org (1/2)
@lifeinboring Depends if you see tomatoes as fruits or vegetables. Anyway, would go for a fruit since there are way more sugar in Ketchup than anything else #healthy
So I think during all these years I underestimated my computer. Adobe Premiere uses 120 Go / 16 Go of my RAM.
(Now I know it relies on the memory of the SSD, but please, optimization seems even worse than FIFA 22)