Content pfp
Content
@
0 reply
0 recast
0 reaction

downshift pfp
downshift
@downshift.eth
as a former infra/ops/noc guy, i’m still very skeptical of AI being able to make building and running software systems trivial anytime soon i’m probably wrong in a way i don’t expect. AI-created and run software may be a completely different paradigm and set of abstractions
7 replies
0 recast
16 reactions

Ryan J. Shaw pfp
Ryan J. Shaw
@rjs
Same. Probably the 16 years in IT consulting for large corporates had made me too cynical, but I just don't see it being commoditized - it'll be a tool that will boost software engineering throughput significantly, but will never make it trivial
2 replies
1 recast
3 reactions

downshift pfp
downshift
@downshift.eth
cc @jpren.eth @bli.eth
1 reply
0 recast
1 reaction

Brian Li 🍊👾 pfp
Brian Li 🍊👾
@bli.eth
I'm fairly confident most of our workflow will be completely erased by AI and on a 3-5 year timespan. The area with the biggest gaps right now are understanding complex systems and exhaustive edge-case coverage when those systems interact with each other. But I could easily see the development of a series of isolated and battle-tested microservices that are then orchestrated in some way with AI. A few months ago, I was building a task manager project for a hackathon with a database and my friend forced me to instead use the context window of the LLM and write it down to JSON blobs that were stored locally. Ever since then, I've challenged my assumptions a bit more about what AI will or will not be able to do.
1 reply
0 recast
2 reactions