Because although AI systems are currently still "hallucinating" incorrect facts in their answers, they are already delivering remarkable results in tasks such as translation, classification, creative writing and generating code.
The hallucination problem is really a big deal that is going to limit AI from dominating any field that requires precision and dependable results.
However, hallucinations are perfect for areas which do not require such, like creative tasks, advertising etc.
I think AI will change how programmers do their jobs, not eliminate them. Messing around with ChatGPT, I was able to write a MUD server much faster than I could from scratch. But I still had to understand the code, tweak some thing.
I think of it as a tool to write a 100k line project in the time it takes to write a 10k line one. I think the potential will be fewer bugs and more productive programmers. That could mean less programmers. Equally likely is software can be much more complex, yet still manageable.
Even if AI replaces software engineers in 50 years, so what? People can move on to other fields of engineering.
2 comments
[ 2.9 ms ] story [ 12.5 ms ] threadThe hallucination problem is really a big deal that is going to limit AI from dominating any field that requires precision and dependable results. However, hallucinations are perfect for areas which do not require such, like creative tasks, advertising etc.
I think of it as a tool to write a 100k line project in the time it takes to write a 10k line one. I think the potential will be fewer bugs and more productive programmers. That could mean less programmers. Equally likely is software can be much more complex, yet still manageable.
Even if AI replaces software engineers in 50 years, so what? People can move on to other fields of engineering.