For this Friday we’re looking at the issue of AI safety. That AI in some form will become a part of banking’s business model, it is worth looking at the work underway to address concerns and real issues in today’s AI landscape.
1. Laissez-faire vs. Stop Everything
Investor Albert Wenger writes a fairly comprehensive research and footnoted analysis of current AI Safety philosophies, including the inherent issues with open-source:
Jeffrey Ladish et. al. demonstrating how cheap it is to remove safety training from an open source model with published weights. This type of research clearly points out that as open source models become rapidly more powerful they can be leveraged for very bad things and that it continues to be super easy to strip away the safeguards that people claim can be built into open source models.
This is a real problem. And people like myself, who have strongly favored permissionless innovation, would do well to acknowledge it and figure out how to deal with it. I have a proposal for how to do that below.
2. US and Europe design agreement
A look at the ‘secure by design’ agreement between the US, Britain, and 16 other countries.
In a 20-page document unveiled Sunday, the 18 countries agreed that companies designing and using AI need to develop and deploy it in a way that keeps customers and the wider public safe from misuse.
The agreement is non-binding and carries mostly general recommendations such as monitoring AI systems for abuse, protecting data from tampering and vetting software suppliers.
Still, the director of the U.S. Cybersecurity and Infrastructure Security Agency, Jen Easterly, said it was important that so many countries put their names to the idea that AI systems needed to put safety first.
3. Biden’s point man on AI
More on the policy side, Politico has a profile of Bruce Reed, White House deputy chief of staff, who is leading the administration’s AI policy:
Reed is a surprising instrument of this transformation. For much of his career, he’s been aligned with a moderate faction of the Democratic Party that tended to befriend Big Tech and distance itself from the anti-business rhetoric common in more liberal corners. In the ‘90s, he was an architect of controversial centrist Clinton administration policies, including welfare reform and the 1994 crime bill.
Now, at 63, Reed finds himself on the same side as many of his longtime skeptics as he has become a tough-on-tech crusader, in favor of a massive assertion of government power against business.
Also a link to the executive order on Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence
And so ends this Friday. For the inquisitive minds, the risks of werewolf astronauts. Click below to let us know how we did: