As you may be aware, there has been a ton of stories about ChatGPT in the news recently. For those who don’t know it’s basically a chat application back by AI. What sets it apart from normal chat app that we would typically build on a website for customer service is the fact that it mimics real life conversation. Not only is a great at conversation but it is great at doing certain tasks for you. An example would be to ask it to write a Python script that imports an excel spreadsheet, sorts the data alphabetically in ascending order, And then rename the file.
As you can probably already see it did not take very long until students started using it to cheat on their assignments. Lately there seems to be a lot of polarizing viewpoints on the subject of AI. Some people love the idea of AI and how it can make us more efficient that our jobs as well as being more productive in our everyday lives. But the latter seems to becoming more reality as AI seems to be taking jobs away from people. You don’t have to look far to see a lot of layoffs from different industries. Is this the tip of the iceberg and will we see more and more companies restructure and downsize because one person can do more with AI than a team of employees?
I prefer to keep an optimistic mind about the future of AI and I see it as a tool. There is one thing I have been thinking about as I design cloud architectures… 🤔 It’s easy to ask AI how to do something. It’s great at answering the “how” questions. But, I find it difficult for AI to answer the “why” questions. It’s in these questions we truly understand the context of why we are even doing a certain task or project in the first place. Without the “why” questions, you don’t really have a clear “how” question. When I look at it from this viewpoint I have to ask myself “Why am I designing this architecture this way?”, “What is the goal of the end-user?” But it’s not only the goal, It’s every step and progression along the way.
It’s asking “Why do we need failover routing for this architecture?” Or “Why should we deploy this application using a blue – green deployment?”, “Should I use a canary deployment instead?”. These might seem like simple questions at a high level but we need to dive deeper. What exactly are the requirements for this specific client? Do they not have the bandwidth for large amounts of data transfers? Maybe they want to mimic the same workflow that they are accustomed to inside their data center and the fully manage cloud solution might not seamlessly mesh with that workflow. What if you are using multiple cloud providers?
It’s going to be within these idiosyncrasies the reason why we need people. To truly understand the client’s needs and their story. This is why I believe AI should be looked at as a tool to become more productive versus a replacement.
What do you think? Will it help boost productivity for you? Or is it bad?
No responses yet