The past few months have been about AI, but all hell broke loose one Friday when DeepSeek was released (unleashed) to the world. NVIDIA stocks were the prime casualty, losing around 20% in a day, because investors got spooked about the hundreds of millions of dollars that now possibly will not be spent on training models that China (says) trained on a mere 6 million dollars.
Of course, whether that figure is true is anyone’s guess, but what does it matter. Markets run on sentiment and if Gamestop taught us anything, it is that rationality is the first casuality in an exchange of greed-based efforts. Which a market is really. There is an occasional Warren Buffet or Jhunjunwala – the rest are busy trying to ‘beat the market’ rather than find long-lasting value.
They say the first ones to go will be the white-collared ones. AI does investment analysis much better than any human (or humans) can. And this will only get better with time. A traditional investment analyst scrutinizes a pool of stocks and makes a decision on some data pointers, coupled with his experience. AI can analyze the whole market, add thousands of data pointers, and come up with recommendations before one can say INVEST. Where does that leave Mr. Analyst?
Oh there is the human factor, judgement call, etc etc arguments.
Humans do the same thing – look at inputs, analyze potential scenarios using experience and learning, and throw out an output. Seems like a workflow to me. Judgement calls may also be made on gut feelings, which are nothing but signals that we take in but don’t understand – stuff like body language for instance, and in financial markets – a hunch. But these too are just inputs to a processor at the end of the day.
Those who still have doubts can GPT their way until realization dawns.
The others – learn to work with your hands. AI beats humans hands down when it comes to computation and reasoning, but still sucks at manual work like dish washing. Your maid’s job is secure (for now), but is yours?
DeepSeek about it.