Yeah, state leaking is definitely starting to show up with some of these tools A nasty one is that files you upload to ChatGPT are available across multiple sessions where you might expect a blank slate, and I think that exposes them to conversations with GPTs too (unless fixed)
TOOLS
-
Discussion on the state of AI models and development tools
By
–
Loads of new model releases but not many actual features & products lately. Should I cover more tools for AI development in the meantime? If yes, which tools are you using? Anyway, I have a Business trip to do next week. Gonna have some time to plan it better.
-
HTTP Libraries vs Custom Request Logic for Developers
By
–
Yeah I should probably take that option more seriously myself – I tend to be happiest writing my own HTTP request calling logic, but if they satisfy developers who prefer a library I should spend time properly figuring them out
-
API Providers Must Ship Client Libraries for Control
By
–
On a related strategic note though: API providers do need to ship client libraries, because if you don't someone else will – and then you could end up in a situation where lots of people are using a client library that you can't release new versions of when you have API updates
-
Python Library Complexity for Direct HTTP Calls to Gemini
By
–
I know Python really well and I still don't like to have to reverse engineer their Python libraries to figure out how to make a direct HTTP call! I had to do that for Gemini just last week
-
Client Libraries Breaking Changes: Trust Issues in API Wrappers
By
–
… which is a bit unintuitive because one would expect that part of the promise of client libraries would be insulating their users from breaking changes made to the JSON APIs that they wrap I have been burned enough times now: I don't trust vendor client libraries not to break
-
API Documentation: JSON Specs Over Client Libraries
By
–
Request for model API providers: please put JSON API docs (with example curl commands) somewhere prominent I always look for those first, before client libraries, because my experience is that client libraries are more likely to have backwards incompatible breaking changes
-
Sweagent vs Devin: Speed Trade-offs and Feature Gaps
By
–
i did end up trying sweagent. its definitely faster but also very very workflow specific to making pr’s for issues. needs a lot more work to have the iterative feedback flow of devin. opendevin seems to be closer, but doesnt have browsing yet https://t.co/NXXSggMdyi
— swyx 🐣 (@swyx) 14 avril 2024i did end up trying sweagent. its definitely faster but also very very workflow specific to making pr’s for issues. needs a lot more work to have the iterative feedback flow of devin. opendevin seems to be closer, but doesnt have browsing yet
-
End to End Latency vs Opus Weaviate Embeddings
By
–
whats end to end latency? vs just opus + weaviate + embeddings
-
Devin AI: Autonomous Task Assignment Over Raw Speed
By
–
On Devin: I have early access (I take no money from them or any other AI lab), and I find it really fascinating, if not there yet. I am not a coder, so I can’t speak to coding. But it is a new method for working with AI, one built around assigning work autonomously, not speed.