AI, the other direction
In order for AI to replace people, all the AI companies need more people to train their models. How ironic.
The latest in that effort is the Mercury project run by OpenAI. OpenAI hired more than 100 ex-investment bankers from the Wall Street’s big banks - JPMorgan, Morgan Stanley and Goldman Sachs. Is OpenAI getting ready for a buying spree now that it became a for profit organization and going public? Maybe that's in the cards, but these bankers are here to train ChatGPT on the 'bankery' stuff.
These people are asked to write prompts and build financial models which are used during M&A transactions, restructurings and IPOs. The objective? To augment or replace junior bankers. As you can imagine, that's bonkers.
The never ending delusional idea, which is still propagated through the media, that somehow you can suck out every single piece of knowledge from people's brain, train a computer on it and get rid of the person or actually the entire class of people. In this case - the junior investment banker.
When you talk to any senior banker who went through the grind, they will tell you that you have to work 80-100 hours per week, weekends, nights. You work under huge pressure to get things done on-time and it has to be absolutely accurate. Mistakes are not tolerated. You do the same work over and over; and it is exhausting. But when you get to the other side, you can spot any mistake in a moment, you can focus on the other parts of the transaction and you become useful to your firm and to your customers. Why? Because they can rely on you and they trust your work. You put your 10,000 hours in.
It sounds so great on paper - we can hire less juniors, they can produce more and the boring stuff will be done for them. Try to extrapolate this to let's say doctors. Why would you expose the newly minted doctors to the grind of residency when they can run everything and anything through Dr. ChatGPT. Will they learn anything? Perhaps how to talk to a computer, but at the end they will become senior doctors by age and not by experience. Next time you are on the surgical table ask yourself which doctor you want to operate on you?
And despite all the billions invested into ChatGPT we still get the disclaimer at the bottom “ChatGPT can make mistakes. Check important info.”
A similar type of lunacy is coming out from Amazon's home security devices division. Its VP of product declared that "We're going to promote based on AI. We're going to promote based on how you're integrating AI into your job." And "The promotion process is the only real incentive that I can give to a team, because that's how they're going to get paid." That's an excellent idea. The message is clear - "I don't care if you create anything useful and valuable for the customer and the company. I only care if you automate your job as much as possible."
One thing I can tell you is that people working in large organizations have truly mastered playing these games. In no time and thanks to AI, there will be templates with charts, bar graphs, describing how much time people saved using AI, how many processes they automated, the percentage of AI utilization, etc. New weekly meetings will emerge where people will sit through presentations given by outside AI consultants about the best use of AI. There will be AI retreats where the best ones will be accepted into the AI Presidential Club and for sure somebody will come up with the 'AI of the week geek' award with a $10 card to Starbucks.
Somehow the questions about strategy, customers and new products will become this obnoxious detail which nobody wants to talk about. Why? It is not what you get promoted for.
I want to make sure that I don't come across as the AI hater, despite my aversion to the term 'AI' itself. My regular readers (I very much appreciate every single one of you!) who follow my writing, you know that I try to separate the hype from reality.
Do I use 'AI'? Of course I do, but I am trying to use it in ways where I end up knowing more about a topic than I did before I used the tool.
Here is an example for you.
For some time, I have been interested in understanding the supply chain of semiconductors, namely processors. Since Nvidia touched the $5 trillion valuation mark, it is hard to escape the talk about AI Chips. But do you know which companies are involved in the making of AI Chips? Do you know what materials are used? What are the steps in making these AI Chips?
Side note: I wrote about Nvidia reaching the $1 trillion valuation in May, 2023.
We are at a decision point in history. We can choose to remove humans from the loop of anything we need to know. Or we can build technology which will make everyone smarter – and better off.
If you want to go down this second path, I have something for you.
I've created this 'AI tool' or, better said, a place, where I can dump any information I find that is related to the AI Chips topic. And I keep asking questions. And I go deep and I go wide. Every so often, the computer says "I don't know". At that moment I start searching for additional information. Once I find it, I upload it there and keep asking questions. Is my objective to replace anyone with this? Of course not. Am I going to be promoted because I am using 'AI' or because I found - for me - a better way to learn about any topic?
Why would one use a tool like this? Perhaps to find where the next innovation might happen and to find promising companies. Are you a VC or investor? How are you building your internal knowledge base about the industry you are investing in? Are you a CEO? How can you spot trends and opportunities? Or if you are a journalist and need to write a well researched article about the AI Chips industry?
The AI Chips are just an example. As you can imagine, you can have your own personal knowledge base on any topic.
This is in my opinion where the 'AI' industry should go. To become the helpful tool which makes you smarter and helps you to find connections to connect the dots. It should not be the threat to make you obsolete.
The recurrent pattern? Smart people with brains never have to worry about their future.