As ChatGPT’s first anniversary approaches, the gifts are coming for the great language model that shook the world. From President Joe Biden comes an oversized “Executive Order on the Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence.” And British Prime Minister Rishi Sunak through a party with a cool theme about the extinction of the human race, wrapped in a Agreement of 28 countries (considering the EU as one country) promising international cooperation to develop AI responsibly. Happy birthday!
Before anyone gets excited, let’s remember that this has been more than half a century since credible studies predicted disastrous climate change. Now that water is literally lapping beneath our feet and the heat is rendering swaths of civilization uninhabitable, the international order has made little or no dent in the gigatons of fossil fuel carbon dioxide released into the atmosphere. The United States has just installed a climate denier as second in line to the presidency. Will AI regulation progress better?
There are reasons to think so. Unlike on the climate issue, where a multi-billion dollar industry has launched an all-out campaign to discredit the threats and thwart necessary measures to reduce carbon emissions, the big AI powers appear to be begging for regulation. They surely have their own interests at heart, but at least there is recognition that rules are necessary. Additionally, unlike climate, governments are taking threats to AI seriously relatively early in the technology’s development. Both the Biden plan and the international agreement represent serious and laudable efforts to deal with AI before it attacks us.
Given this, it seems almost trivial to nitpick about the actual content. But I will do it anyway. Let’s start with Biden’s executive order. I’ve read all 19,811 words of the government’s speech, so you won’t have to. By the end, I was craving Dramamine. How does the president plan to encourage the benefits of AI while taming its dark side? By triggering a human wave of bureaucracy. The document unnecessarily calls for the creation of new committees, task forces, boards and task forces. There is also a constant call to add AI oversight to the duties of current civil servants and political appointees.
Among the elements missing from the document is strong legal support for any regulations and mandates that may result from the plan: executive orders are often overturned by the courts or superseded by Congress, which is considering its own regulation of AI . (Don’t hold your breath, though, because a government shutdown looms.) And many of Biden’s solutions depend on the self-regulation of the industry under review — whose major powers contributed substantially to the initiative.
You can’t fault Biden’s order for its lack of breadth. Almost every AI hot button is addressed in some way, if only to make a wish to come up with solutions later. (This is how he manages the delicate problem of Generative AI and copyright.) Overall, it’s an astonishing commitment to mobilizing the government bureaucracy to tackle all the worrying aspects of a new class of technology, including those most of us have never thought about . Paragraph after subparagraph, the White House orders complex multi-agency studies, each involving extensive industry interaction and expert consultation. Biden’s order tasks bureaucrats with producing complicated reports as casually as some people order DoorDash meals.