Small recap.
Just 5 years back we were in the middle of Covid. Lots of lives were lost. People were in lockdown for many months and doom scrolling. Yet, somehow, humanity gathered itself, organised, not only invented a vaccine but also magically got it delivered to billions of humans around the world. We survived and some of us working in tech actually thrived.
Covid put a lot of digital tech in focus. Lot of money was thrown around. Tech became super hot. And then it melted.
Offline took over again. People realised the valuations were way out of propotions. Meanwhile, interest rates started climbing and put pressure on margins.
So, there was a correction and we transitioned quitely from “Work From Home Forever”, “The Great Resignation”, “Quiet Quitting” etc to layoffs and “Back To Office - Whether you like it or not”.
But something was brewing on the side which everyone first ignored , then laughed about and then it became so good that you hope it doesn’t fight you.
Yes, the AIs, the LLMs have gotten so good that if you aren’t using them OR building on them, then you are going to be on the wrong side.
But let’s cut to the chase.
LLMs are forcing us to deeply introspect the nature of our work and what it is we really do. If a task can be relegated to an LLM, it will be.
There are several good reads on this from AI as a copilot to the end of programming altogether to a more far reaching view.
Things are still unfolding. Right now atleast, it feels like a calm before the storm, or the next big AI model. It maybe that AI scaling has hit a wall.
For now, the focus has shifted to applications Generative AI’s Act Two.
One of the interesting ways to look at it is through the lens of humans “in the loop”, “on the loop” and “off the loop”.
AI as a co-pilot is going to grow. It can become more effective, more personalised as a “Pair Programmer” than even a human. That is going to have far reaching consequences to jobs as we know it.
Entry level SWE are already feeling the heat.
If we look at how autopilot tech on planes change pilot training and skill requirements is any indication, knowledge work that needs a human in the loop will also change similarly with AI copilots.
https://chatgpt.com/share/6702bc9c-844c-8012-90bf-08f43e00c39f
While autopilot has changed pilot training by shifting some focus to system management and human-machine interaction, the fundamentals of flying—manual control, situational awareness, and emergency response—remain critical. The role of pilots has evolved into more of a systems manager, but their responsibility for the safety of the flight is as crucial as ever. Autopilot is a tool, not a replacement, and pilot training ensures that aviators are ready to handle both automated and manual flight situations effectively.
And that can’t be a bad thing. Right?