Andrej Karpathy: Software in the era of AI [video]

I easily see a huge future for agentic assistance in the enterprise, but I struggle mightily to see how many IT leaders would accept the output code of something like a menugen app as production-viable.

Additionally, if you're licensing code from external vendors who've built their own products at least partly through LLM-driven superpowers, how do you have faith that they know how things work and won't inadvertently break something they don't know how to fix? This goes for niche tools (like Clerk, or Polar.sh or similar) as much as for big heavy things (like a CRM or ERP).

I was on the CEO track about ten years ago and left it for a new career in big tech, and I don't envy the folks currently trying to figure out the future of safe, secure IT in the enterprise.

Put another way, when I cause bugs, they are often glaring (more typos, fewer logic mistakes). Plus, as the author it's often straightforward to debug since you already have a deep sense for how the code works - you lived through it.

So far, using LLMs has downgraded my productivity. The bugs LLMs introduce are often subtle logical errors, yet "working" code. These errors are especially hard to debug when you didn't write the code yourself — now you have to learn the code as if you wrote it anyway.

I also find it more stressful deploying LLM code. I know in my bones how carefully I write code, due to a decade of roughly "one non critical bug per 10k lines" that keeps me asleep at night. The quality of LLM code can be quite chaotic.

That said, I'm not holding my breath. I expect this to all flip someday, with an LLM becoming a better and more stable coder than I am, so I guess I will keep working with them to make sure I'm proficient when that day comes.