Book Review: I, Human: Tomas Chamorro-Premuzic
Is humanity done, or has it just begun?
I was with Tomas Chamorro-Premuzic this week talking about his new Harvard Business Review book, ‘I, Human’, which reflects on the impact of AI (and broader technology) on human behaviour. I did think of asking GPT-4 to do a multimodal presentation on the event, but actually that wouldn’t support the spirit of the book (and we’re already seeing far too many of those posts anyway), so this article accentuates the human side of the event instead.
It was my first trip back to an HBR event since the pandemic, and the first time seeing Tomas since his talk on ‘Incompetent Men’ at the same venue four years ago. It was great meeting him again, but even nicer to catch up with other people I know better but whom I’ve not seen for a few years, like Sian Harrington from the People Space. But the highlight has to be seeing Natalie Cooper who I used to know when she worked at Personnel Today and the Changeboard, and was very supportive of early HR bloggers like Scott McArthur and myself, and who I’ve not seen for well over ten years. It was great meeting others, especially Jennifer Bryan, Sanjay Bhogaita and Nikita Mikhailov, for the first time too
I knew at the time that this early period of social media development (around 2007-12) was a very special time, mainly because I could anticipate the impact that even this early evolution of digital technology would have. But also, and what I hadn’t appreciated was, that as social media expanded, it’d get a bit less special too. So I still love meeting my Linkedin connections today, but it’s perhaps a little less meaningful than meeting people I used to know from blogging or Twitter etc back then.
I also didn’t understand that social media would get corrupted too and turn into the hate filled cesspit that so much of it is today!
We talked about this over drinks after the event and I asked Tomas if he thought social media had somehow acquired humanity’s dark traits over the last twenty years, and his view was that it absolutely has. (Although Jennifer’s challenge that social media can still do a lot of good, such as in the Me Too movement, is still very valid too.)
I mention all of this because I promised you I’d take a human perspective in this article. But also because it’s social media which lies behind today’s hyper-connected world which then enables the ‘datafication of you’ and has led to the development of AI prediction machines. And because Tomas’ book suggests the above experience is likely to be repeated as we progress through the current evolution of digital technology with GPT-4, other AI and broader automation. Unless, of course, we make sure it doesn’t.
However, based on Thomas’ book and presentation, that may be even more difficult this time. AI works when we automate ourselves, making ourselves more predictable, and working pro bono to help algorithms reverse engineer our decisions:
“When we buy a product Amazon has recommended to us, watch a movie Netflix has suggested, or listen to a playlist Spotify has curated for us, we are making data-driven changes to our life, conforming to an algorithmic syntax that eliminates the behavioural differences between us and people like us, boosting the valuation of tech firms by making our livesmoer predicatble. Prediction improves through tow different ways: the algorithms get smarter of the humans get dumber.”
This unleashes even more of our dark traits and squeezes out our humanity, for example by making self enhancement an expected behaviour whereas it would be seen as obnoxious in the office. It has also never having been easier to be misinformed, even though knowledge is now ubiquitous. And our addiction to digital devices means that even if productivity while working on tasks goes up, in terms of total time it goes down, because so much time is wasted.
Tomas suggests there are pockets of optimism in all of this which are about harnessing the opportunities where AI won’t learn, and using skills it won’t master, for example, empathy, self awareness and curiosity (though what about artificial empathy systems like Cogito?). And using these to differentiate ourselves from machines, and other humans.
Also debiasing the workplace by being ethical by design. Tomas suggests that if we focus on these sufficiently, then generative AI shouldn’t degenerate humanity!
Personally, I’m not sure that what’s happened in social media over the last twenty years can give us much confidence that we’re up to the task.
I’ve not yet read Reid Hoffman, and GPT-4’s ‘Impromptu’ on amplifying our humanity through AI, which I will do. However, at first glance, it looks to be quite focused on spending even more time in helping AI learn, like all those Linkedin carousels sharing people’s favourite prompts. So yes, it absolutely offers opportunities for increased productivity, apart from all the time we need to put into sustaining it.
Thanks to Tomas for the book, presentation, and wine, to Sally, HBR and Havas for the very human event, Nikita for more wine, and everyone else I met, for some fulfilling discussion, and suggesting the great sub-title for this article!
Director, Strategic HR Academy