The title of this article has been a common title in the past decade. About two years ago, in a team building keynote speech, I told my team the most important mission of a tech leader is to prepare the transformation of the department to those transformations which time finally arrived.
My vision at that time was we will build software 2.0 solutions which credit goes to Andrej Karpati. We will more on describing a need with words and provide test data instead of the actual program code writing.
With the arrival of ChatGPT and especially the GPT 4.0 model now, it has become a reality. I feel the same empowerment as with the arrival of the Internet when I was more successful than most of my colleagues because I was able to use the search with proper keywords and relations.
Now we have the same experience with Artificial Intelligence, but we not only access the information, but AI condenses the most relevant information in summary.
I can imagine how much volume will increase in demand for prompt engineers who can ask the right question from the AI!
We just simply rub ChatGPT and he fullfill our wishes. Like in the fairy tale it can be used for good and bad things, but now the ghost is out from the bottle.
Now we are over the inflection point, and the pace of the changes become exponential. Those who can leverage person/business these tools will create more significant value than the competitors who not using.
In the past two weeks I peer worked with the GPT4 ChatGPT, and I can tell, my productivity significantly increased, even on fields where I never worked.
I had one problem solving challenge where I noticed with simple google searches I hardly find the solution, however with these tools, I had the answer in minutes.
I also using it for programming, however in most cases these are boiler plate codes. Even with these tasks it is extremely helpful.
Just imagine you have to create an interface for an online service? How much time you need until you go through the full documentation?
Instead you can ask a code in the programming language which you are familiar, and have the function what you require.
The next stage will be absolutely the one, where we have some on-premise intelligence which can be augmented by external services such ChatGPT. This way we can protect our IPs but leverage the knowledge of the outside world.
The biggest challenge in these tools will be the same as in a self-organized organization culture. You must have trust in your processes and colleagues!
Sooner you gain this skill, sooner you can step into the next stage which become the true Software 2.0. Where you bake a specific image to your hardware including all the codes from firmwares and os, processes and application functions.
These systems will 10-100x more productive than the ones we are using today!
But at this point we will have no chance to understand in the whole how it works. We will use data and textual descriptions do define requirements and systems.
With the baked codes we can eliminate many duplications and redundancy of each layer, and because of these efficiency factors even the most significant businesses can be served with on-premise resources. This way our data can be owned by truly by ourselves, while we can leverage cloud based cognitive services for the generic code baking.
These systems will create different requirements in the IT sector, most of the actual roles will transform to new roles, such the prompt engineers.
I'm looking forward how far we go in the next half year in this journey!