Their idea was that OpenAI was so far ahead of the competition no one could ever catch up. Turns out they weren’t and now they’re at the bottom.
this post was submitted on 18 May 2026
954 points (98.5% liked)
Programmer Humor
31506 readers
1543 users here now
Welcome to Programmer Humor!
This is a place where you can post jokes, memes, humor, etc. related to programming!
For sharing awful code theres also Programming Horror.
Rules
- Keep content in english
- No advertisements
- Posts must be related to programming or programmer topics
founded 2 years ago
MODERATORS
If agentic LLMs are the prize, I’d be fine with losing.
I mean, Microsoft’s biggest mistake was shoving it in front of everyone’s faces. The real reason that all the other “agentic BS” is received well is because the people who use them have an actual use case, or, are very enthusiastic about the technology and enjoy messing with it. Thus the discussion is mostly from that small group of people who will have something positive to say.
The truth is, that all the models and harnesses suck for most use cases that most people have. When you shove it in front of a general audience and make them interact with it, then the discussion will be about how bad it is.
It's weird because copilot in office tries to push agents on you as if it were a Jehovah's witness.
So GitHub copilot doesn't have them? I don't really use that.
It's saying Copilot was the first on the scene and had access to literally all of the training data anyone could possibly want, and is still being shown up by most other AI models. Their failure to capture the vibe coding space is a legendary fumble. At least that was my read.
Eh wait. Copilot (any of the about 30 products with copilot in the name) is not a model. Microsoft makes a few models like phi but they're underwhelming. All of copilot runs on models from external parties like openai and anthropic. So basically Microsoft is at the mercy of their own competitors. They're in the awkward position that providing training data to their model providers not only improves their own product but their competitors' as well.
Additionally, Microsoft's most profitable market is enterprise and they would absolutely shiver at their data being used for training and would abandon the service in droves.
Despite being "all in on AI" Microsoft is in a really vulnerable position. Their added value is their integration with their other services (and data therein through RAG).
load more comments
(3 replies)
load more comments
(2 replies)
I see this as proof of how bad LLMs actually are. You have an AI trained on essentially humanity's collective programming library. Languages of machines and computers. The result should be ungodly and near perfection. If there was any semblance of understanding in AI, it should be revealed in it's capability to produce code.
Although... I can definitely see Microsoft thinking that their code is the example of perfection and training copilot on that rather than github.
should be ungodly and near perfection
Counterpoint: garbage in, garbage out
load more comments
(1 replies)
Your proof of how bad LLMs are is the fact that there are a bunch of other companies producing way better coding agents and coding models than Microsoft is? I'm not sure how that follows. Those other agents are good, that's the point of this.
load more comments
(6 replies)
load more comments
(1 replies)
Microsoft could have been king with with chatgpt for personal superapp, github copilot for developers and something like sharepoint/power vibe widgets. But nooo, they make windows recall when ai models can't run locally
if they had made a unified copilot agent they would have won. Think open claw with the power of NPUs for small tasks and cloud for big queries dedicated APIs for interacting with all the microsoft products special tailored version for developers. More focus on retrieving information and doing small tasks for the user than generating slop.
The first versions they released were so fucking bad and every app had basically just a chatbot with zero functionality. It ruined the product for when it could actually do tasks.
I was using tabnine before copilot was even announced... they are the ones who fumbled this more than github.