📚 Finished reading: A Thousand Brains by Jeff Hawkins, having no idea how it ended up in my Kindle library. I am glad to have opened it, as I now have some semblance of a framework for how this thing we call intelligence might work. Note that the newest developments in neuroscience are just a starting point, as most of the book deals with their implications for AI and the future of humanity. If that sounds like overreach, know that by the end of the book it is. Still, these wafer-thin speculations don’t detract from the book’s meatier parts.
Confirmation bias alert: the framework repeats almost word for word the thought I had a while back — and more recently — about AI, that true general intelligence needs to be able to interact with its environment. So I may be blind to some obvious deficiencies in the argument. But then again, great minds, etc.
I will have more to write about this soon (ha!), but until the stars align for an extended writing session here is a good opinion piece from FT’s John Thornhill about why LLMs may not be all that great for lay people dabbling in, for example, medicine:
When the test scenarios were entered directly into the AI models, the chatbots correctly identified the conditions in 94.9 per cent of cases. However, the participants did far worse: they provided incomplete information and the chatbots often misinterpreted their prompts, resulting in the success rate dropping to just 34.5 per cent. The technological capabilities of these models did not change but the human inputs did, leading to very different outputs.
The emphasis is mine, because it is a neat summarization of what I wrote 2 years ago. Humans are unique not because of what’s inside our heads but because of how we interact with the environment. There will be no artificial general intelligence until that problem is solved.
Some of the best blog posts are rants, and Andrew Gelman just published one, about reckless disregard for the truth. Here is why he thinks the term “bullshit” does not apply:
In my post, I asked what do you call it when someone is lying but they’re doing it in such a socially-acceptable way that nobody ever calls them on it? Some commenters suggested the term “bullshit,” but that didn’t quite seem right to me, because these people seemed pretty deliberate in their factual misstatements.
I disagree. Whether the bullshitter is deliberate should not matter, and many do indeed BS with a specific goal in mind. In the examples he lists those are inflating the impact of a paper and getting paid for expert testimony in favor of big tobacco. Indeed, dig deep enough and you will find hunger for money and prestige to be at the root of much bullshit.
An excellent blog post that is not a rant: Single-function devices in the world of the everything machine, by Christopher Butler.
Limitations expand our experience by engaging our imagination. Unlimited options arrest our imagination by capturing us in the experience of choice. One, I firmly believe, is necessary for creativity, while the other is its opiate. Generally speaking, we don’t need more features. We need more focus.
Indeed.
The NYT is collecting readers' votes for their list of 100 best movies of the 21st century and, well, I made some interesting choices. I do stand behind each and every one of them!
For those not using X, I picked:
A few good links to start the week:
Thomas Basbøll is back writing, with a wonderfully meta-post about why one would want to write at all:
The obvious alternative that I’m heading towards is to seek reasons to write within yourself, rather than in your environment. Write for the clarity it brings or the pleasure it affords. Write because it improves your mind, not the minds of your readers. In the future, as most of the prose we need to get by (the prose that stores and transmits useful information) is produced by machines, we will write for the same reason that we swim, rides bikes, jog, go to the gym.
That is the dream.
Never have I ever seen a civilian car racing down the National Mall like it’s their own back yard… until now. Summer heat makes people do crazy things.
Fact of the day from FT’s Edward Luce:
America has 120 guns per 100 people against 4.6 in England and Wales. The next highest democracy to America is Montenegro with 39 guns per 100 people — still a third of its level.
Super-excited about our family’s Montenegro vacation in August.
I have mentioned before that I am not a fan of IQ as a measure of anything other than un-intelligence and have linked to Taleb’s short essay on it from way back in 2019. Well, that same year Sean McClure wrote an even more thorough account of why testing intelligence as done today is pseudoscience, and you get to learn much more about models, biases, and the scientific method. Recommended long read.