@7ang0n1n3@GithubProjects Why be so judgemental? Who cares if he used Codex, Pi, Claude, etc? Did he ask your opinion? ... Oh wait, this is the Internet, your opinion must be heard by all, continue on... for a moment, I forgot where I was.
If you vibecode, this has probably happened to you: You step away for five minutes and come back to uncommitted work trashed because the agent ran git stash to "clean up" or git checkout . Happens too often in my monorepo with 5+ agents running, so I wrote this. (Details below)
I built an AI that roasts GitHub repos.
https://t.co/WaRCY9U6M2 got a B-: "NetlifyTab.tsx is so large it has its own ZIP code and congressional representative"
An Uber clone got an F.
Paste any GitHub URL β https://t.co/ZoSLtxuQfJ
#vibecoding#buildinpublic
An agent I'm working on used Bing to search `type 2 diabetes solution competitors` and the results it found....well, I thought it hallucinated the results, so I went to Bing and typed the query manually. Bing...bing bing bing...what in the world are you thinking, Bing? π€¦π€¦π€¦
@fernandorojo Wooot woot!! That is incredibly awesome! If there's any help needed, I would be interested to try and help. Just as a learning experience as well! I know firsthand how tricky keyboard management can be in mobile apps so deeply empathize with the work you guys have done!
@pamir_ai Is there any official support for using opencode with the beautiful integrations that you've already done like the LED? And can you confirm if we can get root access to the device?
@PetSafe do you have any doors taller than 91" or so? Looking for something like this: https://t.co/ZKomkqWJGe - But my door track is 113.5" tall! Can you help?
@pseudomoaner @gm8xx8 Yeah I am so not a fan of trying to shove reasoning into a black box like this. We need reasoning to be done external to the llms. Not internal. I think this black box approach is going the wrong way in general with llms. They're great at generating language, they do not think.
1 in every 7 dollars is spent on treating diabetes globally thatβs $700B! Diabetes care now evolves with salesforce Agentforce. Precina leverages these AI agents and data cloud and apps to fill critical gaps in patient care, analyzing patient data to personalize diabetes management plans, providing support where current systems fall short, and ensuring care reaches underserved populations. By optimizing resource allocation and scaling access, Agentforce empowers lives, reduces costs, and transforms diabetes care into a safer, more compassionate system. β€οΈπ€
@how_many_roads_@thomaspeters@Benioff@agentforce Have you looked at how AgentForce works with data cloud and how Customer 360 has been re-engineered to enable the metadata access from the LLMS? Sounds like they've already resolved your complaint. Now if the customer data is bad to begin with, SF can't fix bad entry ...
@thomaspeters@Benioff@agentforce 90% of all companies i supported in their Salesforce journey had horrible customer data. Agentforces value has yet to be seen on scale - my hunch, it will take a while and sentiment has its just another SF gimmick like NFT Cloud was
@Benioff@agentforce It's not really about any breakthrough technology, is it? It's all about the integration, the fact that Salesforce has everything so deeply integrated together is what drives the difference. Agentforce is truly more than the sum of its parts! β¨ Great messaging too, @Benioff πͺπ₯
@ollama@huggingface Yeah you still don't support log probabilities in your API though? When are you going to actually support log probs? That would be far more exciting than just running unhugging face