One of my favourite Disney films is The Jungle Book. There are some real bangers in there, as the kids would say: The Bare Necessities and I Wanna Be Like You being the standouts.
But I was put in mind of another song recently when Donald Trump and a menace of leaders from tech and AI companies met to announce the Joint Commitment on Frontier Responsibilities. Yes, I did use the collective term of ‘menace’ for tech and AI leaders here because sometimes it feels like that’s what they are.
Anyway, back to Mowgli and his escapades in the jungle and the scene that came to mind when I read that the menace had signed a “morally binding” agreement to put controls on AI. Any guesses? It’s when Kaa, the python, sings the song Trust in Me while hypnotising Mowgli so that he can eat him later.
At the announcement of the agreement, Trump claimed “it’s almost like a constitution in a way, and the biggest people in the world signed that, and I signed it as president, and it really is a form of protection”. In light of all we know about his own “morality” and Trump’s pick and mix approach to which parts of the constitution are inviable and which should be replaced with something more pleasing to him, this does not fill me with great confidence.
Trump also claimed the agreement would enable the companies to engage in “tremendous self-policing” of their products and technologies. Again, given what we know about the tremendous self-policing these guys (it’s nearly all guys, isn’t it?) have applied to their products and technologies over the years, this fills me with exactly the same level of confidence as Trump’s comparison of the agreement to the constitution.
Trust us, bro
I don’t think I’m being unfair here. I stand to be corrected but I would be amazed if, for example, technology companies were cited in business schools or other such institutions, as definitive examples of best practice for self-policing. It does seem a wee bit bizarre that at a time when the warnings around the development of AI are becoming increasingly apocalyptic, the response from Trump and the technology and AI leaders, who are pretty much to blame for the situation that we find ourselves in, is “Don’t worry, we’ve got this. Trust us, we’ll keep you safe”.
Trust. It’s a small word but it has a very high value. It carries a lot of weight. The thing about trust is that, for the most part, it doesn’t happen in isolation. You don’t trust someone completely out of the blue. It builds based on past experience. Past experience does not inspire trust in this case.
Usually, if there’s little or no trust, you seek to buttress any agreement with safeguards and regulations. Not in this case. All of the controls it contains are voluntary.
As The Guardian notes: “None of the controls listed in the agreement involve government regulators or include a commitment to publicly detail the findings of independent evaluations. The agreement also appears to allow companies to pick their own evaluators, appoint their own oversight boards and decide whether to publish their results.”
I’m trying to imagine what will happen if this agreement becomes the de facto approach to the development of AI and also what would happen if every other industry on earth had a similar self-policing arrangement with zero government regulation or enforcement. It’s so far beyond my comprehension and capabilities that I’d need help to try and visualise it.
Luckily, Trump and the menace are working hard to provide that assistance with their development of ‘super intelligence’. The same super intelligence that, when given its proper name of artificial intelligence, presents a growing threat to our continued existence. But it’s fine. They’ve asked us to trust them. Who are we to argue?
Ltd
