The Evolution
I’m from the times when, as an Automation Test Engineer, I was looked at with a red eye by manual testers — for automating everything and unintentionally giving them an impression of “automation is going to take away all the jobs of manual testers”. However, I was getting praised by rest all of them for saving their time and energy, by automating most of the tests and now there were faster feedbacks for the backend developers. Which means they won’t be bugged anymore in the middle of their next project for a bug caught in the previous project they worked on. These were the times we used Waterfall model and not agile.
Then came cloud automation. Infrastructure for the cloud can be created through scripts — that’s wow factor. Terraform working like magic was hard to resist to learn and also to understand what are the trade-offs here? how do we handle scalability without blowing up the cost? etc. DevOps felt enjoyable with the cloud infra automation.
An advanced version of automating things is what ChatGPT seemed like. For something that started as a chatbot where — you talk to it about your problem and it hands you out the solution and involves no human it feels like an ultimate automation. It doesn’t stop there, it evolves further to go for Agentic AI where each AI agent, uses the set of models and tasks and delegates which agent should do what based on how we configure it. If I think I need to be involved, I can configure human-in-the-loop. If I don’t think I need to be bothered about this anymore while I work on something else, that is handled too. By now we have a lot of models, lot of agents, and lots and lots of vibe coders too.
We are now in the age of “don’t tell me you can’t build” era. There are apps that can build more apps just through conversation. It would create your app, test it for you and fix issues and vulnerabilities too to some extent. It’s totally your call if you want to rely on that when it comes to scalability which brings it’s own set of issues. I saw a couple of apps and labelled them mentally as “good enough for a hackathon, may be”. Now please don’t judge me! I’m not being all skeptical about it. It’s just that the SDET in me somewhere is alive who keeps asking “are you sure you want to sign that off?” each time I see some of those apps, even if I’ve transitioned to an AI Data Engineer role now.
How does the product perform at scale? How does it impact the cost? Will the infra scale down when needed? How secure is this app? — these questions are too hard to let go for someone with a Quality Engineering background.
The Scaling Data
For a moment, let’s forget the technicalities and focus on use cases that are harmless. Think of an AI generated content for entertainment where people will forgive you if you generate a video of a person talking to you and suddenly you see that the person has a 3rd ear that shows up and disappears , an audio generated based on what I typed out so that I don’t need to give it my voice, or a video generated for a story I wrote which I think would be loved by kids and helps them visualize it well, a huge book summarized into 10 pages of data or further summarized to say 1 page of it.
Now scale this, as in, imagine thousands of people doing this in a given hour. Can you imagine the amount of data out there?
What I mean to say is, it’s astonishing to see that if I have an original piece of 1 page information about something I love being written and put out on the internet, there is a possibility of that being accessed by the models ( lots of them) and generate pages together of information and put back on the internet again for the same loop to continue. This means, the amount of data we have around us, probably comes from a small book or a small page of original source and there could be 100s of content on the same info , all linked together through hashtags.
Analysis Paralysis
With all this abundance in data, comes a side effect of the same perk and this term is described by wikipedia as
“Analysis Paralysis — a state of overthinking and overanalyzing a situation to the point where a decision is never made, halting all forward progress. It stems from a fear of making the wrong choice, information overload, or a desire for a perfectly risk-free outcome.”
Have you found yourself in this phase where you have a lot of information and it feels like you have a lot more to catch up but still, every point both the argument and the counter argument feels sane in your head and they are conveniently stopping you from going further. Congratulations! You now understand that there is a term for it.
Skimming the information from what is offered to us is a present day skill. The sooner you master it, the sooner you will save yourself from falling into the loop of constant questioning for simple decisions you would have made earlier even at the cost of risking the outcome. Taking risk feels very intimidating , I understand! but are you at peace without taking it? This question is for you to answer Well skimmed and consumed data could be gold. It just needs to be channelled to the right question.
Access to super powers
We all have our own super powers that we are extremely good at something but then there are things we don’t call ourselves as an expert. But that doesn’t have to stop you or your ideas anymore. To bring an idea to life, tools like Claude help you with everything you need and ask you if that’s what you wanted or do you need further changes.
This comes from a personal experience where I wanted to build a website that spoke about my profile and career trajectory, but I didn’t have enough time for it and honestly that was just an idea and not an interest that I wanted to invest my time in.
Enter Claude Code, I had to tell what I wanted, how , which details to include and which not to, and I got my website with the steps to host it freely. I even was asked if I need to make changes to it or I have better ideas to improvise them. Trust me, on asking this question, the things I mentioned as improvements were not even a part of my initial thought process and all of this was within an hour on a weekend! I’ve written more about it here for you to read further.
So where does all this leave us? We are living in the times where tools are ready and the barrier between idea and possibilities is thinner than ever. I don’t know who will replace whom but the human stays human — validating, asking questions, making decisions . The SDET in me will always ask “Are you sure?” each time there is something new, “Let’s do this!” when confident and I’ve made peace with that. Keeping the balance is the key, re-inventing the skills is the way forward and the trick is to not get confused.
