Weizenbaum’s nightmares: how the inventor of the first chatbot turned against AI

This is the best summary I could come up with:

His father, Jechiel, who had emigrated to Germany from Galicia, which spanned what is now south-eastern Poland and western Ukraine, at the age of 12, was an accomplished furrier who had acquired a comfortable foothold in society, a nice apartment, and a much younger Viennese wife (Weizenbaum’s mother).

The programs that ran on MIT’s system included those for sending messages from one user to another (a precursor of email), editing text (early word processing) and searching a database with 15,000 journal articles (a primitive JSTOR).

“And once I started thinking along those lines, I couldn’t stop.” In the last years of his life, he would reflect on his politicisation during the 1960s as a return to the social consciousness of his leftist days in Detroit and his experiences in Nazi Germany: “I stayed true to who I was,” he told the German writer Gunna Wendt.

A glimpse at the index reveals the range of Weizenbaum’s interlocutors: not only colleagues like Minsky and McCarthy but the political philosopher Hannah Arendt, the critical theorist Max Horkheimer, and the experimental playwright Eugène Ionesco.

Certain of Weizenbaum’s nightmares have come true: so-called risk assessment instruments are being used by judges across the US to make crucial decisions about bail, sentencing, parole and probation, while AI-powered chatbots are routinely touted as an automated alternative to seeing a human therapist.

For instance, experts such as the linguist Emily M Bender draw attention to how large language models of the kind that sit beneath ChatGPT can echo regressive viewpoints, like racism and sexism, because they are trained on data drawn from the internet.

I'm a bot and I'm open source!

It is long but a very good article. He seemed like an interesting person.