AI, Twitter, and Misinformation

Generating misinformation is easier, cheaper, and faster than ever. Anyone can exploit AI systems: from conventional deepfakes to language models like GPT-3 and text-to-image diffusion models like Stable Diffusion. These AIs can’t tell right from wrong and often make up realities that have no grounding in the real world. At the same time, because these systems evolve so fast, we’re lagging behind in detection and recognition countermeasures.

And I’d add here another problem on top of this. Recommender systems, which literally decide what we see on Twitter (or any other social media app), are also driven by opaque AI algorithms. This wouldn’t be much of a problem if it wasn’t for Musk’s recent decisions: first, he decided to remove identity verification, and then he fired the team in charge of algorithmic responsibility.

Just like the dilemma faced by humans in selecting a restaurant, these [AI] agents also struggle with balancing the time spent discovering better actions (exploration) and the time spent taking actions that led to high rewards in the past (exploitation). Too much curiosity can distract the agent from making good decisions and too little means the agent will never discover good decisions.

If we expect digital assistants to facilitate personal tasks that involve a mix of public and private data, we’ll need the technology to provide “perfect secrecy,” or the highest possible level of privacy, in certain situations. Until now, prior methods either have ignored the privacy question or provided weaker privacy guarantees.