Estelle Platini

'#AI is a marketing term, not a technical term of art. The term “artificial intelligence” was coined in 1956 by cognitive and computer scientist John McCarthy – about a decade after the first proto-neural network architectures were created. In subsequent interviews McCarthy is very clear about why he invented the term. First, he didn’t want to include the mathematician and philosopher Norbert Wiener in a workshop he was hosting that summer. You see, Wiener had already coined the term “cybernetics,” under whose umbrella the field was then organized. McCarthy wanted to create his own field, not to contribute to Norbert’s – which is how you become the “father” instead of a dutiful disciple. […] Secondly, McCarthy wanted grant money. And he thought the phrase “artificial intelligence” was catchy enough to attract such funding from the US government'.

Meredith Whittaker acceptance speech: helmut-schmidt.de/en/news-1/de

#metrics #probabilities #ModelCalibration #MachineLearning #ML #ethics

The Prizewinner's Speech

In her speech, Meredith Whittaker warns against the…

www.helmut-schmidt.de
Estelle Platini

Calibrating models to obtain acceptable predictions

The sources said that the approval to automatically adopt #Lavender’s kill lists, which had previously been used only as an auxiliary tool, was granted about two weeks into the war, after intelligence personnel “manually” checked the accuracy of a random sample of several hundred targets selected by the #AI system. When that sample found that Lavender’s results had reached 90 percent accuracy in identifying an individual’s affiliation with Hamas, the army authorized the sweeping use of the system. From that moment, if Lavender decided an individual was a militant in Hamas, the sources were essentially asked to treat that as an order.

“Still, I found them more ethical than the targets that we bombed just for ‘deterrence’ — highrises that are evacuated and toppled just to cause destruction.”

972mag.com/lavender-ai-israeli @israel @data

#metrics #probabilities #usability #ModelCalibration #MachineLearning #ML #OutputAudit #FoundationalModels