Data Disquiet

As large language models’ popularity grows, so does concern about their tendency to “hallucinate” or generate information that is factually incorrect or even made up, errors caused by problematic data sets or incorrect assumptions made by the model. The questionable results produced by chatbots have led to growing disquiet among users, developers and policy makers. Current piecemeal approaches to data governance do not reflect these models’ complexity or the magnitude of the data upon which they are based.

In this new paper, Susan Ariel Aaronson argues that policy makers need to develop a systemic approach to address these concerns, and recommends incentivizing greater transparency and accountability around data-set development.

Argentina’s economy is in a state of crisis. President Javier Milei, who took office on December 10, inherited a bankrupt government, an overvalued peso and a spaghetti bowl of informal exchange rates flourishing under the watch of a reserve-stripped central bank. Milei won the fall elections by tapping into voters’ weariness and vowing to use a chainsaw to cut government spending.

Hector Torres writes that “Milei has made sweeping promises and raised expectations. He has the courage to reprimand the country’s political caste. But to make Argentina ‘great again,’ Milei will also need to learn to build consensus.”

Apr. 25 – 3:00 p.m. EDT (UTC–04:00): Please note new start time.
Join us online for a discussion on ⿻數位 Plurality with E. Glen Weyl and Paul Samson.

While digital technology threatens to tear our free and open societies apart through polarization, inequality, loneliness and fear, on a delicate, diverse and politically divided East Asian island things are different. In the decade since the occupation of Taiwan’s parliament, this island of resilience achieved inclusive, technology-fuelled growth, and entrusted the people to tackle shared challenges such as environmental protection, while capitalizing on a culture of innovation to “hack the government.”

Tickets are free; find out more and register here.

“If there’s one thing American presidential hopefuls will agree on this year, it’s that the country needs to get ‘tough on China’…. Last year’s updated export controls on advanced computer chips, the Biden administration’s executive order to boost the US biotechnology industry, and large investments in automated, uncrewed weapons systems were all designed with this supposed new Cold War — and the fear it might turn hot — in mind.”

But Stephen Clare and Christian Ruhl say the same capabilities that make these technologies important also make them dangerous: “Technology competition today cries out for new thinking about foreign policy that considers global risks and remains open to cooperation with rivals on critical issues.”