I use AI daily, I have access to all of the best models through work, and part of my job is to see how they integrate into our existing tools. I sleep easy every night knowing that an AI absolutely cannot write a python script to summarize their financial outlook for the next quarter. There is absolutely no way it will give anything remotely close to accurate, and it is doubly impossible for it to give you a script that will work next quarter. Now, I know what you’re thinking; “that sounds like a fairly manageable task if given the data, most other tasks would require human interactions and more work”. And my answer to you is: “Yes.”
In fairness, the robot in the comic passes the turing test by spouting agile terminology and empty promises, much like some developers I’ve worked with.
I’m sure that any rational corporation understands this perfectly and is more than willing to keep paying their employees instead of fall for the hype and the dream of a (seemingly) perfect tool just to end up with a bugged product needed to be heavily fixed later.
I use AI daily, I have access to all of the best models through work, and part of my job is to see how they integrate into our existing tools. I sleep easy every night knowing that an AI absolutely cannot write a python script to summarize their financial outlook for the next quarter. There is absolutely no way it will give anything remotely close to accurate, and it is doubly impossible for it to give you a script that will work next quarter. Now, I know what you’re thinking; “that sounds like a fairly manageable task if given the data, most other tasks would require human interactions and more work”. And my answer to you is: “Yes.”
In fairness, the robot in the comic passes the turing test by spouting agile terminology and empty promises, much like some developers I’ve worked with.
I’m sure that any rational corporation understands this perfectly and is more than willing to keep paying their employees instead of fall for the hype and the dream of a (seemingly) perfect tool just to end up with a bugged product needed to be heavily fixed later.
Radical I’m not reading that! 💥 🦵
Why do you use the llms and instead not use them and lie and say you did but get accurate output?