4 months ago
I love this definition of an AI agent today – “Brilliant, but lacks expertise.” This is 100% how I see AI agents today. I found this video interesting, if only because it gave a non-hyped version of agents. You wouldn’t hire a developer to manage your network, and that’s the same way you should approach the development of your agents – one size does not fit all.
4 months ago
Problems are easy to identify, trickier to articulate. “The tire is flat.” Why? I don’t know, but it’s flat, maybe it has something to do with driving over those jagged pieces of wood? The clear articulation of problems is once again becoming the defining trait for software developers; we’re finally moving from knowledge of languages and into “figuring out what is wrong and finding a solution for it.” But it starts with figuring out the problem so everyone understands.
4 months ago
You’re probably getting assigned tasks that are above your knowledge. Good, that’s the idea. The next question is, what are you going to do about it? Are you going to level up to the task or stay where you are, waiting for the ones that line up with your skills to come along? If you’re waiting for the tasks you’re suited for to come along, you’re missing the narrative; those tasks are going away, they are diminishing, they are over. The only option is to level up. In another way, think of a video game. When you level up your
5 months ago
AIs promise is to save time for the good things in life. To do that, we need to give it carefully crafted prompts and assignments for it to do its work. In some cases, we pay AI to do this work for us, and we hope AI is up to the task. The thing is, AI isn’t up to the task just yet; it’s getting there, but the fulfillment of the promise is not going to happen for a while. So what do you do until that time? Work the problem, don’t throw it in the garbage, but work the
5 months ago
Do you prefer this answer or this one? Do you want me to keep going? Do you want me to try again? No, be like a student, handing in their paper – you get one chance – give me your best. Be methodical, be precise, be clear and simple. People who fail multiple times aren’t doing so because their work was sloppy; it’s because it hasn’t yet, but failing because you threw some random things together and hoped it would succeed – that’s not good work. That’s being lazy.
