Notices by mmitchell_ai (mmitchell_ai@mastodon.social)

"Meanwhile, a practical step in limiting emissions would be to not integrate AI into platforms that don’t need it, said @huggingface 's @SashaMTL ."
https://www.wsj.com/articles/artificial-intelligence-can-make-companies-greener-but-it-also-guzzles-energy-7c7b678

“Fundamentally speaking, if you do want to save the planet with AI, you have to consider also the environmental footprint [of AI first],” she says. “It doesn’t make sense to burn a forest and then use AI to track deforestation.”- @SashaMTL ❤️🔥❤️🔥 https://www.theguardian.com/technology/2023/aug/01/techscape-environment-cost-ai-artificial-intelligence

6 (!) years ago, @timnitGebru put forward a profound idea for AI research: document your training data. To say she was treated like shit would be putting it lightly. If she had been respected like her male peers were, imagine what a different path AI could have taken.

@onlybrownmastodon I'm always open to being linked to by @Mer__edith !!

Okay, #60Minutes is saying that Google's Bard model "spoke in a foreign language it was never trained to know." I looked into what this can mean, and it appears to be a lie. Here's the evidence, curious what others found. ?

1-The exact language referred to seems to be Bengali.
https://dailymail.co.uk/news/article-11980033/Google-CEO-says-doesnt-fully-understand-new-AI-program-Bard-works.html
I can't stomach the 60 minutes piece; here's the Daily Mail naming it. (Will check transcript when avail.)

Study showing that ChatGPT can influence our moral judgments, and we underestimate how influenced we are. Highlights the need for education to help people better understand what these systems are doing.
https://www.nature.com/articles/s41598-023-31341-0