16 Followers
33 Following
29 Posts
Here for technology & development chat | Urbanism | Work in AI @IBMResearch
Europe
Went to see the elephants at a temple festival at Tripunitara. Nobody seems to know when the temple was built. The brilliant percussion performance lasted 90 minutes. Slow repetitive start, gradually increasing tempo and polyrhythmic sections that had that crowd going. Surreal.
Google Maps in India now shows isochrones for drive times. Nice little nugget. Would be interesting if they did public transport as well.

Nice paper on "blue zones" - areas of the world where people have remarkable longevity.

Paper suggests fraud and error, rather than some lifestyle choice.

https://www.biorxiv.org/content/10.1101/704080v2.full

Fingal County Council's noise map is a neat little resource. Bathed in din we are in D15.

https://fingalcoco.maps.arcgis.com/apps/webappviewer/index.html?id=8e3bb95ccef04d34b6dfa040676106a5

This car crash of a paper compares humans to ChatGPT based on carbon emissions associated with writing one page of text.

If only lifecycle analysis were so easy...

https://arxiv.org/abs/2303.06219

The Carbon Emissions of Writing and Illustrating Are Lower for AI than for Humans

As AI systems proliferate, their greenhouse gas emissions are an increasingly important concern for human societies. We analyze the emissions of several AI systems (ChatGPT, BLOOM, DALL-E2, Midjourney) relative to those of humans completing the same tasks. We find that an AI writing a page of text emits 130 to 1500 times less CO2e than a human doing so. Similarly, an AI creating an image emits 310 to 2900 times less. Emissions analysis do not account for social impacts such as professional displacement, legality, and rebound effects. In addition, AI is not a substitute for all human tasks. Nevertheless, at present, the use of AI holds the potential to carry out several major activities at much lower emission levels than can humans.

arXiv.org
Wow, didn’t expect Kentaro Toyama to be quoted at #icml2023 who is hugely critical of techno-solutionism.
Google’s medical AI chatbot is already being tested in hospitals

Research hospital Mayo Clinic has been among those testing Google’s new medical chatbot, Med-PaLM 2, since April. Google’s research shows doctors prefer its answers over those of doctors, though they can be less accurate.

The Verge
I've heard about this, but my first direct experience of a paper which states that data is supposedly available - but not really when you ask.
Fascinating (and a bit crazy) story of Kazimierz Nowak, a Polish traveler, who cycled the length of Africa twice in the 1930s.
The ChatGPT system card is worth a read IMHO and comes with a content warning. 😂 (while the API itself does not).