Great and diverse content this week on cyber threat actors Russia and China, the use of AI voices in campaign ads, city level planning for AI, and how we might make the internet less toxic.
We were thrilled to host the authors of "Battlefield Cyber: How China and Russia are Undermining Our Democracy and National Security" at Pitt Cyber this week. Their book clearly lays out the ways in which coordinated cyber-attacks against U.S. public and private sector entities poses a threat to national security, identify the reasons it's been so difficult to mount a coordinated response and offer recommendations for an effective offensive strategy.
This research by Grow Progress poses a fascinating question: what is the effectiveness of using AI-generated voiceovers for political ads? With AI all but certain to be used in future campaigns, research into questions like this is critical.
Case Study: Can AI Voices Persuade Voters | Grow Progress
An interesting illustration AI bias, with the model ignoring the prompt instructions in deferring to the built in bias of the training data.
AI was asked to create images of Black African docs treating white kids. How'd it go? | NPR
NYC has released a city-level plan for government adoption and regulation of AI. Some great actionable items here, we'll be watching this with interest.
The New York City Artificial Intelligence Plan
A fascinating analysis pointing to the free structure of the internet and the resulting turn to targeted advertising as the source of all the toxic things about the internet: the "epidemic of sadness, of loneliness, of meanness, that seems to self-reinforce in many online spaces." In a recommendation reminiscent of this episode of Wired's 'Have a Nice Future' podcast, the author forecasts a move towards a more decentralized internet, enable by federation.