AI tools are close to being able to do a host of dangerous things, like walking people through the mistakes they made during a failed attempt at making dangerous pathogens and guiding them to a better protocol. There needs to be a durable way to probe the ways these systems might be misused, even as newer and more powerful technologies are continuously released.

The people building the next iteration of AI technology are growing concerned with how lifelike the next generation of generative content has already become.

In an interview with Axios, an unnamed “leading AI architect” said that in private tests, experts can no longer tell whether AI-generated imagery is real or fake, which nobody expected to be possible this soon.

As the report continues, AI insiders expect this kind of technology to be available for anyone to use or purchase in 2024 — even as social media companies are weakening their disinformation policies and slashing the departments that work to enforce them.

Computer science professor Fernanda Viégas believes “dashboards” that disclose AI models’ biases can help laypeople control AI.

How many years will pass before transformative AI is built? Three people who have thought about this question a lot are Ajeya Cotra from Open Philanthropy, Daniel Kokotajlo from OpenAI and Ege Erdil from Epoch. Despite each spending hundreds of hours investigating this question, they still still disagree substantially about the relevant timescales. For instance, here are their median timelines for one operationalization of transformative AI:

You can see the strength of their disagreements in the graphs below, where they give very different probability distributions over two questions relating to AGI development.

So I invited them to have a conversation about where their disagreements lie, sitting down for 3 hours to have a written dialogue. You can read the discussion below, which I personally found quite valuable.