The State of Tomorrow

We live at a crossroads. On the one hand, we have all the promises for humanity, because of scientific progress and innovation happening on a daily basis. On the other hand, we have all these people painting a doomsday picture. Which one is right? Both viewpoints are valid, but the question is what do you lean into? Let's turn a leaf of history to understand this dilemma. When the industrial revolution happened, a lot of people thought that many jobs would be lost. And they were right. But many jobs were created. The best part was that our productivity went up. We became process-oriented. When the internet came in the 90s, it was ushered as the greatest revolution of our times. Then came the mobile phones. Today we can't live without one. Today, we are at such a point again. The AI revolution is tantamount to the discovery of fire.

But this time it's different. The fire or the assembly line had nothing to do with intelligence. Now, we are trying to wield it, without understanding it ourselves. There are 16 dimensions to our mind, of which, intelligence is just one of them. Most of the mind dichotomy can be subsumed under two broad categories. Intelligence and Memory. Our emphasis on intelligence and memory has shaped our educational systems and workplace. We appreciate knowledge and reward the person who has more knowledge than others. While knowledge is good to have, it's not mandatory, as we have made it be. A machine trained on the whole of the internet has much more knowledge than one single person. That is just the beginning. We have taught machines to infer this knowledge and become intelligent. The story doesn't end there.

Machines are breaking barriers to what once was thought to be solely human characteristics. Today they are creative and can solve problems. GPT-Next is coming in October and it's purported to be 100 times faster than its predecessor. Text translation is a solved problem. So is text to speech and images. Text to video would be ready in a year or two. As we exhaust the contents of the internet, synthetic data is powering the next breed of LLMs. Llama (from Meta) is an open-source model that has been downloaded by 300 million people. Claude (Anthropic) and Grok (X.AI ) are raising the bar. The way we do programming has changed. Earlier it was INPUT + CODE = OUTPUT, Now it's, INPUT + OUTPUT = CODE. Models are getting better and more powerful.

The printing press could not write a book. The assembly line could not correct itself. Gadgets couldn't take decisions by themselves. Now, it's different. Self-driving cars make decisions every second without human intervention. An AI radiologist can spot tumor where a human expert could not. Smart chatbots have begun replacing humans in the call center. This new breed of intelligence is not just good at pattern recognition, but self-learning and correction. More and more of our decisions are getting outsourced to AI. In a decade's time this new species will make us complements or redundants, in some case. We are in a strange dilemma. Without knowing ourselves in our entirety, we are trying to understand a new species and its impact on our future. Can we just pause the speed a little bit or make a correction? Too late. The train has already left the station.

God Bless!

((Rajesh Menon))