July 30th newsletter
Thought for the week: what's your passion? How can you be brave enough to live it? (Don't make this only about work, work isn't everything)
I continue to be astonished by what you can do with AI. Just last week I built a visual basic application, a desktop application, and programmed a AI assistant to download some files for me. Here's the kicker. I don't know visual basic or really know python other than just rudimentary edits or translation.
I wouldn't call this exactly vibe coding. I do have some experience in coding how to get a computer to do math and math is ultimately a step by step application of rules to solve problems. Just like coding.
I don't think someone is truly a vibe coder could actually get much out of the coding applications without having some understanding of the holes you can fall into while coding.
I’ve found that all of the ai tools have a tendency to fall into ruts and not be able to get out of them and having someone with some expertise in how to navigate out of ruts can help it get unstuck.
There were multiple times that I was trying to get the AI to fix something and it just didn't seem to be able to and I was able to use my understanding of how to code to correct it. These conditions may not exist forever. I mean, obviously these tools are improving. But I continue to believe that the best situation for using AI is a situation where you have an experienced person working in tandem with the technology. Even if it makes a lot of mistakes or doesn't do things perfectly, I continue to be astonished by what you can do with these tools, especially if you have some expertise in order to guide it or notice when it's failing to do what you ask.
I've seen people refer to AI users's cyborgs and I actually think that's an apt analogy. It's not so much that the AI can operate independently within experienced people or that the AI can do things on its own because it's ultimately kind of stupid even if it has an enormous amount of capability. The the best case is a highly skilled person working in tandem with the tool.
Most importantly, there is a learning curve, you have to understand how to recognize when it's failing and when it's not going to be able to do what you asked and knowing when you step in is something that has to be learned.
Like all technology, it's often hard to determine if it's magic or what exactly is happening.
You can see why people that are experienced often make a lot of mistakes with these tools because they don't know how to recognize the tool is making mistakes.
But if you are a skilled professional in your field and you're not using these tools to make you better, I just don't know what you're doing at this point.