In his September 2024 essay, ‘The Subprime AI Crisis’, Edward Zitron said the following (emphasis mine):
“I am deeply concerned that this entire industry is built on sand. Large Language Models at the scale of ChatGPT, Claude, Gemini and Llama are unsustainable, and do not appear to have a path to profitability due to the compute-intensive nature of generative AI. Training them necessitates spending hundreds of millions — if not billions — of dollars, and requires such a large amount of training data that these companies have effectively stolen from millions of artists and writers and hoped they’d get away with it…My concern is that I believe we’re in the midst of a subprime AI crisis, where thousands of companies have integrated generative AI at prices that are far from stable, and even further from profitable.“
For people who use ChatGPT, Claude and/or Gemini, what would you do if the subscription prices of these LLMs became so prohibitive?
Where would you draw the line and opt out? $30/month? $100? $500? Never even bought a subscription to start with?
Everyone is riding the AI wave. Yet, no matter how high the waves rise, there’s always a non-zero probability that they will come crashing down at some point.
So, how should you protect yourself? Cultivate these two things: (1) domain expertise, (2) taste.
Here’s the thing: LLMs might be better and faster than humans at prediction and pattern recognition. But when it comes to making judgments about what should be valued, we humans really shine.
Unfortunately, without domain expertise and a cultivated taste for what is good, your value judgements are essentially useless.
As it stands, we humans are not born with domain expertise. And the only things we are biologically hardwired to taste are sweet, bitter, sour and salty, not ideas and insight. Moreso, no one develops domain expertise by osmosis, neither can anybody cultivate taste just by having someone else lay hands on our heads.
Both domain expertise and taste must be earned via blood and sweat and tears and time.
Which is why it is never a bad idea to read the classics. Read within your field. Read outside your field. Have conversations with different people. Travel widely as you’re able to. Build things. Start projects. Cultivate new experiences. Do something tangible in the real, physical world. Write about your learnings and share them.
When you do these things, you just might cultivate the domain expertise and taste that will make you indispensable both during the current LLM boom and in the post-Gen AI world.
I’m on the same journey with you. And that’s one other reason I have made the recommitment to share my unpolished thoughts more.