[ad_1]
In letter Deepmind and the European Institute of Bioinformatics have published a database of more than 350,000 3D protein structures predicted by the artificial intelligence model AlphaFold.
These data cover the roughly 20,000 proteins made in the human body and are available for anyone to study. Proteomes from 20 other organisms, from zebrafish to E. coli bacteria, are also found there, and hundreds of millions of additional structures will be added over time, we are told.
“In the hands of scientists around the world, this new protein almanac will enable and accelerate research that will advance our understanding of these building blocks of life,” said Demis Hassabis, CEO of DeepMind. He hopes it will be a valuable resource that will be used in the discovery of new drugs and our understanding of diseases.
It all depends on the accuracy of the AlphaFold’s predictions, of course. It dominated the CASP protein folding competition last year. It is about 95% accurate in determining the locations of individual atoms in the protein with an accuracy of one ngstrom for the best studied proteins, we are told.
Investors Still Donate Controversial Clearview AI Facial Recognition Millions of Dollars
Clearview AI, the controversial newcomer that has received multiple cease and desist requests from tech companies, has secured $ 30 million in Series B funding from investors.
Best known for taking three billion photos from public social media accounts to develop face matching software, Clearview is often used as a case study of why unregulated use of technology can be harmful.
He is under investigation by data and privacy watchdogs in the UK and Australia, and Canada has cracked down on the upstart.
Talk to the Dead via GPT-3
Here is the gripping story of a man who, overwhelmed with grief by the death of his fiancée, decided to polish a GPT-3 based chatbot to mimic his personality so he could speak to her beyond the grave.
Joshua Barbeau discovered Project December last year, a website that allows anyone to create a chatbot, trained on provided snippets, that writes in a distinct way. The app uses OpenAI’s GPT-2 and GPT-3 text generation models in its backend. People type in a sentence and the machine responds.
Out of curiosity, Barbeau handed him old texts and Facebook messages that his fiancée, Jessica Pereria, sent him during their relationship. While he knew it was make-believe and the chatbot would occasionally say things that were unlike Jessica, there were instances in their conversation that seemed surprisingly very real.
It is interesting to see how he used the chatbot to feel better about his death. It’s worth reading to find out what he did knowing he could spend a limited time talking to him before he was deleted. ®
[ad_2]
Source link