Want to HIRE us to implement AI into your Business or Workflow? Fill out this work form: td730kenue7.typeform.com/to/WndMD5l7 💗 Thank you so much for watching guys! I would highly appreciate it if you subscribe (turn on notifcation bell), like, and comment what else you want to see! 📆 Book a 1-On-1 Consulting Call WIth Me: calendly.com/worldzofai/ai-consulting-call-1 🔥 Become a Patron (Private Discord): patreon.com/WorldofAi 🧠 Follow me on Twitter: twitter.com/intheworldofai Love y'all and have an amazing day fellas. Thank you so much guys! Love yall!
single prompting for one shot dev is interesting, but conversational dev and iteration is where I think there is more power. Been playing with Databutton, some real promise there
Hi there, I have a question. If you answer this that would be soo helpful. Is it possible to have a coding agents team in a larger workflow each from different model (like; deepseek, codestral, GPT4o, Sonnet 3.5, etc.) It's like multi model team, each agent from different model working together on same task of the team. There will be multiple teams working together to complete the whole task.
Yes its called mixture of agents. There are several Implementations that are very useful allowing smaller models to match the performance of larger models. Though the examples i saw had pretty big models for local use anyway.
Sure its possible already. For example you can even have them in parallel. For example, you could have both gpt-4o and sonnet-3.5 refiners that return values to the orchestrator that sends it either human in the loop or maybe an AI grader that you code based on whatever criteria you have (AI detection for example). It could choose which is better, or merge them. You might also find that one model performs better at writing code. At some point, we may write agentic code that automatically selects the model based on the specific task and a framework benchmark API.
Anyone looking to sell an AI products good luck. Obsolete within the first 1-3 months! Best chance for a lasting product is one that actually works well with local LLMs - designed from the ground up to make the best of the worst.
[Must Watch]: Claude Enginner: Powerful AI Coding Agent Framework!: ruclips.net/video/-VW8dJbEDsk/видео.htmlsi=YSjk8G0kWHwaxdV5 Gemini-UI-To-Code: AI Develops Frontend - UI Generator (Opensource): ruclips.net/video/MA9ErYq6luo/видео.htmlsi=LTSCbXTRODakpgfA NEW Claude 3.5 Sonnet: The Best LLM Ever! Closest to AGI?: ruclips.net/video/Rt5JVhyeOIw/видео.htmlsi=nT2uf135_8Rps6F4
Want to HIRE us to implement AI into your Business or Workflow? Fill out this work form: td730kenue7.typeform.com/to/WndMD5l7
💗 Thank you so much for watching guys! I would highly appreciate it if you subscribe (turn on notifcation bell), like, and comment what else you want to see!
📆 Book a 1-On-1 Consulting Call WIth Me: calendly.com/worldzofai/ai-consulting-call-1
🔥 Become a Patron (Private Discord): patreon.com/WorldofAi
🧠 Follow me on Twitter: twitter.com/intheworldofai
Love y'all and have an amazing day fellas. Thank you so much guys! Love yall!
single prompting for one shot dev is interesting, but conversational dev and iteration is where I think there is more power. Been playing with Databutton, some real promise there
Hi there, I have a question. If you answer this that would be soo helpful.
Is it possible to have a coding agents team in a larger workflow each from different model (like; deepseek, codestral, GPT4o, Sonnet 3.5, etc.)
It's like multi model team, each agent from different model working together on same task of the team.
There will be multiple teams working together to complete the whole task.
Yes its called mixture of agents. There are several Implementations that are very useful allowing smaller models to match the performance of larger models. Though the examples i saw had pretty big models for local use anyway.
Sure its possible already. For example you can even have them in parallel. For example, you could have both gpt-4o and sonnet-3.5 refiners that return values to the orchestrator that sends it either human in the loop or maybe an AI grader that you code based on whatever criteria you have (AI detection for example). It could choose which is better, or merge them. You might also find that one model performs better at writing code. At some point, we may write agentic code that automatically selects the model based on the specific task and a framework benchmark API.
This is possible. I don’t know if anyone has done it.
Using sub agents it's good and powerful but as of now aider still on top when it comes to simplicity
Anyone looking to sell an AI products good luck. Obsolete within the first 1-3 months!
Best chance for a lasting product is one that actually works well with local LLMs - designed from the ground up to make the best of the worst.
true🥲🥲🥲
Why local llm's?
Gorilla LLM: Writes API Calls BETTER Than GPT-4o? - Signs of AGI: ruclips.net/video/Ac1ZfLgkWHQ/видео.html
[Must Watch]:
Claude Enginner: Powerful AI Coding Agent Framework!: ruclips.net/video/-VW8dJbEDsk/видео.htmlsi=YSjk8G0kWHwaxdV5
Gemini-UI-To-Code: AI Develops Frontend - UI Generator (Opensource): ruclips.net/video/MA9ErYq6luo/видео.htmlsi=LTSCbXTRODakpgfA
NEW Claude 3.5 Sonnet: The Best LLM Ever! Closest to AGI?: ruclips.net/video/Rt5JVhyeOIw/видео.htmlsi=nT2uf135_8Rps6F4
GraphRAG: Ultimate RAG Engine - Semantic Search, Embeddings, Vector Search, & More! - ruclips.net/video/kHZHMzv3Shg/видео.html
Can you create videoa which the project contains without GPT Open AI api rather than some open source and free API key models!!
You can always ask chatGPT that question
try hugging face models or use the gemini api
Aider + Claude 3.5: Develop a Full-stack App Without Writing ANY Code!: ruclips.net/video/dzOWn8TI738/видео.html
Mesop: Google's New AI UI Builder!: ruclips.net/video/BK44wq2itwQ/видео.html