There was a period in my life when I swore I wanted to be a doctor or an attorney. It was because I was sick and tired of my industry, of ‘tech’. I hated the way that digital technology so easily served as the handmaiden of postmodern slop. Even when the internet was new, and then when n-tier architecture was new, and then when the cloud was new and now that social media has gone from new to tiresome there was always something missing. What was missing was the edification of the humanities. In retrospect, many more of us can see the damage done by COVID panic, and by Woke catastrophizing and by the double standards of multiculturalist diversity initiatives. The lies circumnavigated our world before the truth could get its pants on.
Now we have AI, the perfect attention sink that has been trained on all of the word salads of the past two decades of our blather. Oh yeah and every PhD thesis, corporate whitepaper, US patent and the whole of academic publication. It has become the perfect companion to curiosity. I am about as perversely curious as anyone can be. It is the weakness that may destroy me.
I’ve always categorized problems into four sections. Puzzles, mysteries, rabbit holes and black holes. Puzzles have all the pieces for their solutions right in front of you. A mystery is a puzzle with pieces missing that you need to find or create. A rabbit hole is a problem whose solution creates more problems. The worst of these are the black holes which is a rabbit hole you do not realize you are in. Artificial intelligence is the black hole of the mind. You get all the answers you ask for and the initiative and satisfaction of your process of discovery is itself a problem.
AI provides you a pre-digested consensus. You completely avoid the friction of needing to approach, address and comprehend human feedback. The human feedback loop has been shortcut into perfect jello cubes of interaction. You get what you ask for to ‘a reasonable approximation of the truth’. It’s not all good, but it’s all good enough.
Read the benchmarks in the way that I read them. Consider that a score of the AI capability as a percentile. If it gets a B+ then perhaps it’s as smart as 87% of humanity. Why would you ask anyone else?
The Deepest Question
I came to a realization a couple weeks ago that the biggest problem I face with AI is that I don’t know what to ask it. You might recall an advertisement from several years ago when an ISP showed people who had ‘reached the end of the internet’. That old internet, the one conquered by Google, is just a small fraction of everything humanity might know, because now AI is generative. It creates new content. It creates new content for you because you’re now able to ask more questions than the old internet search provided for you. Now you can act and create more content.
Every once in a while, as I’m using Claude on Opus 5 High (the pretty much max) it offers me a term I’ve never heard before. So I ask what that means. Uh oh. Rabbit hole. Now I go on a curiosity tour of discovery and I find somebody who has already found a complex solution to a problem I didn’t know existed ten minutes before. Explain! I command. And I get an answer, a double dose of humiliation. Why didn’t I know this before?
AI, in this way, presents an infinite democracy of knowledge with no resolution at the end of the day. Everybody can know everything within the bounds of the ethics the AI makers have deemed as acceptable for the public. Yet we have no idea of how many people know what, where they are, what they want to do with it. It’s a knowledge economy where the buck stops someplace invisible. They say knowledge is power. I’m not getting any more power, are you?
We’re all living in our own private Idaho.
Nice
AI , for right now, is LLMs. LLM's that can't beat the Atari 2600 in chess.
Until such time as that changes, & it may, I am not interested.
That said, there are some things that it can do.
Regarding that, is your job search helper ready for Beta testing by 3rd parties? I am interested .