I even went in and edited the text area size iirc from 8k to 32k or something just so I could paste longer context into it.
I really felt like an elite haxor.
However, times have changed. What was "state of the art" in 2023 is pedestrian now. Copilot really had an early lead, in my opinion when Bard felt somewhat off. Now? I don't even think about Copilot. I feel very comfortable putting my thoughts in Claude or even Gemini.
ROFL
Might as well be Copilot at this point with how CLIs have been adopted.
I do feel like if any of the major companies could do with a rebranding it would be copilot. They are tossing that name on all of their stuff, and it just doesn't carry the weight of any of the big names even though its chatgpt models under the hood. Personally i associate it with annoying bloatware, and silently judge windows users based on if that icon is still on their tasbar.
I don’t know why I get annoyed when LLM’s and their output are casually referred to as “he/she”, particularly by non-techies, but I do. There’s something about personifying an LLM that seems incorrect. Perhaps it’s a fear being stoked that increasingly, people might actually be thinking of LLM’s as living beings.
I think the costing there is the problem and with (GitHub) Copilot. Not owning their own model and not able to take advantage of the (probably brutally subsidized) fixed monthly packages that have relatively generous limits means Cursor and Copilot can't compete cost wise on a per-token basis.
JetBrains has similar problems with its "Junie" product.
Many some of these will have second chances once Anthropic and OpenAI run out of $$ runway and are forced to charge something closer to actual cost. Or if the Chinese open weight models do some more catching up.
lordleft•58m ago
Configure0251•14m ago