AI Is Revolutionary, But Let's Take a Quick Breath
We need to deal with the inhuman at a human pace
We’re already living in a time where our collective inability to deal with eco-grief — and move past denial to sustainable future — threatens our entire planet. Our first contact with AI is threatening to do the same. AI will radically change our world and we need to come to terms with that. Life without iPhones, social media, or the internet is now almost unimaginable to those under 30, and AI will quickly display change an order of magnitude greater.
Every single piece of digital information that you can send, receive or imagine sending or receiving can now be — or very soon will be — duplicable or fakeable in real time. Your child’s voice, your lover’s video FaceTime, your accountant’s signature, your favourite password all replaced and reused. Whatever tenuous human control we individuals ever had over the internet in the face of the massive tech giant algorithms will dissolve.
However, this is not a piece about the dangers of AI, or even a piece about its promises. There are many places you can go to read the latest thinking about the phenomenon but they are almost instantly outmoded and outdated. That’s because one of the three main significant aspects of AI is its exponential growth. In 6 years it has increased in speed and scope from being a game-winning concept to a game-changing reality.
Imagine an ad studio and a focus group that was able to use billions of existing data points to modify both the efficiency of the focus group as well as the efficiency of the message almost instantly. Imagine how rapidly and effectively it could hone the ad’s message. Every experience that takes humans 10,000 hours to reach mastery will take an AI five minutes.
The second attribute of AI which changes it from just the Amazon or Facebook algorithms on steroids, is its ability to generate language. One of the names for the particular types of AI that we have now is GLLM (pronounced “Golem”) and which stands for Generative Large Language Models. The point about those is that not only that they learn by imitating, but they can also generate human-like language which gives computers new skills and, paired with the ethics-free environment of user-generated content platforms (known, bizarrely, as “social” media), will lead to even further toxicity from Musk, Zuckerberg et al.
The third aspect of AI, which is almost impossible to grasp, is its ability to scale. Until now, algorithms have been specific and, even if vast, have been limited to particular outfits. Even though their properties are largely unknown and, despite some lawsuits, unaccountable, Bing, Google, Amazon are recognizable outlets with proprietary algorithms. The new AIs will transcend those like Google dwarfed Ask Jeeves.
AI is going to take advantage of the ubiquitous availability of personal data that no one ever talks about (credit score data, behavioral data, location data, etc.). This wealth of readily available data will empower autonomous agents to "test and learn" how best to persuade people to do almost anything. We will see countless numbers of companies start using this super-powered ad-tech this year. — Shelly Palmer
There’s real reason to suspect that the AI on Google for example, would eventually “talk” to the AIs on other platforms. In fact, that’s exactly what the AI boom has grown out of, the ability of different fields and different programs to form their output in a way that allows them to talk to one another. At the moment, there’s not a good understanding of how AIs could or would communicate with each other, but it’s a simple emerging property that will inevitably come.
It’s extremely human to understand our dreams and aspirations for the future in terms of what we know now and what we’ve grown up with. But working through our eco grief is helping us shape our understanding of a changing world. Knowing that we will grow older in a world with an irreparably changed climate and whose population has doubled since 1973 (and also doubled between 1950 and 1987) can allow us to take action to save this crowded planet. Knowing that AI will transform our coming decades will help us be a central part of that transformation.
For the moment, AI doesn’t think, it just imitates. And if it does gain some form of self-consciousness we can hope that orthogonality will mean that its superhuman abilities will give it superhuman ethics. But, at a pace much faster than we (or our religions) have dealt with our eco-anxiety we must face up to, and overcome our AI-fear. If we don’t embrace and shape the fact that AI will transform pretty much every aspect of society whether political, economic, cultural it will transform them despite us, rather than for our benefit.
Happy 2023!
This has been my first newsletter post of the year because I’ve barely written for myself. Instead, I have spent much of this year as a small and part time, but busy, part of setting up Shtetl: Haredi Free Press.
It’s a new 501c3 nonprofit news organization providing independent coverage of Ultra Orthodox Jewish communities in the New York area. My friend Larry Cohler-Esses wrote about it for the New York Daily News and I hope to write more about why this particular black hole of news at the heart of one of the world’s great news cities is so particularly fascinating.
I also did write this piece for its launch, explaining the name, “Shtetl.”
Sci-Fi Bylines: Most of the bylines I’ve produced so far this year have been my regular monthly SF reviews for Book and Film Globe (you can listen to them all, too, I think) I think my favourite books have been the new ones from Ann Leckie and Annalee Newitz, but some interesting ones all around.
Climate Justice: Climate is a religious matter. If you want to read about the intersection of Judaism and Eco-anxiety — which is the type of effort I think we should rapidly make to deal with AI — my friend and colleague Rabbi Laura Bellows wrote a wonderful piece about the Spiritual Adaptation workshops she puts on for Dayenu.