The Escalating Cost of Carelessness

Carelessness and AI Are a Lethal Combo

There are many lessons one could take away from Sarah Wynn-Williams’ damning account of her years as public policy director at Facebook/Meta. Careless People is a must-read cautionary tale for people concerned about the future of a world with AI. Wynn-Williams joined Facebook smitten by the awesome promise of technology to connect and empower people in the political process. What she witnessed instead was the damage caused by greed and power amplified by machine learning algorithms.

Possible Take-Aways

• Being more successful/rich/powerful than everyone is not evidence that you are smarter than everyone else.
• The decisions made by leaders who lack moral principles tend to be bad for society at large.
• If you can control the information people see, and the timing and frequency of that information, you can radically change individual and social behavior.
• If you relentlessly prioritize profit and growth over all else, some will get rich, and all else will suffer.

What Meta Wants

Mark Zuckerberg was in the right place at the right time with a good idea - once. His next half-a-dozen ideas were relative flops, including Facebook phones, Internet.org and the billions he spent on the metaverse he renamed his company for, and yet he is still one of the most powerful people in the world. This is the power that technology can confer on the stochastically gifted.

Our last era of technology was the age of information networks. The power of connected digital information created Google, Facebook, Amazon and Twitter among others. These companies and their leaders have had enormous impact on the world - the things we believe, the work we do, the way our institutions function.

The era of machine learning and artificial intelligence is ahead of us, and if our current operating model persists, AI will confer even greater power on an even smaller, luckier group of leaders. We cannot afford for them to care as little for their impact on our collective well-being as Zuck and his enablers did. We need either better-informed collective governance or a greater ability to hold those with outsized power accountable for their actions and decisions.

We should not rely on chance that those in control of AI will be the principled and courageous leaders we need. Though Meta was relatively late to the LLM-party, they are now using their substantial resources to catch-up. Given their past actions it would be naive in the extreme to assume that Meta’s open-source strategy to AI models is altruistic. They want to win. They want to own AI and the awesome power that machine intelligence confers. If we learn one thing from Wynn-Williams courageous book, it should be that what is good for Meta is likely bad for the rest of us.