I am writing thread, essay and story monthly about programming, story-telling and user interface. Building Curioucity and Instill-Ai. My opinions are my own.
This is a persistent thread about the progression of my new novel, it's written in Chinese but I will share the learning and funning thing in English under this thread.
#早餐一顆水煮蛋#AEggForBreakfast
I always feel confused by the resort's trail map
They are 2D, and it's hard for me to understand each slope feel like on the map
So I build a 3D model preview to help me understand the resort more
#japowdb
Previously our culture emphasize the merit of doing and making, and compose it to natural manufacturing level.
When you produce you earn a small pie of the market. And our culture think that is the way, the is life's ultimate form.
But in the era of AI, producing is not that important anymore, we demand more people with hand-on experience of creating stuff. Creating means things that not existing before but you figure it out and show it to the world.
Join us for @hacktoberfest this October! 🎉
If you are fascinated by #AI, #opensource and unsure where to start, this would be a great entry point!!
This year we have exciting new challenges and opportunities to win cool #swag like Amazon Gift Cards worth US$50.
Links below 👇
> Read our latest blog for challenges and how to contribute
> Join the official @hacktoberfest #Discord community to meet other contributors and learn together
Let's make #data AI-ready together!
#InstillAI #Hacktoberfest2024
Seeking better Structured Outputs from LLMs?💡
"Structured Outputs doesn't prevent all model mistakes" - OpenAI
We explored a more efficient & accurate approach using Instill VDP + OpenAI JSON mode.
Check out our findings: https://t.co/rqp66eYOfo
#InstillAI#LLMs#AI
🔔 Discord Office Hour: Today at 3 pm BST (GMT+1)
🔗 https://t.co/xaPzwnLQFj
Come explore 🔮 Instill Core, a full-stack AI infrastructure tool for data, model and pipeline orchestration!
Highlights:
🎼 Orchestrate data, models, and pipelines to focus on fast AI iteration
��� Build AI-powered pipelines for unstructured data ETL up to 10x faster
🔌 Utilize pre-built components for diverse data sources, AI models, and third-party tools
Let's talk about #AI and #Infrastructure!
🧠 AI isn't just about models - it's a full ecosystem.
From data ingestion to pipeline and model orchestration, modularity is key to robust AI systems.
Learn how to unlock AI's true potential in our latest blog:
🔗 https://t.co/G2RHt6yKau
#AI#MachineLearning#MLOps
🚀 Whoa!
Our GitHub project, VDP, has reached 1k stars!
https://t.co/L5GFYOza6j
We couldn't have done it without our incredible community of developers, contributors, and supporters. Thank you for all the support. 🙌
#InstillAI#OpenSource#GitHub#Milestone#1KStars#Community
🌟 Exciting news! Join us for Hacktoberfest 2023 with @illacloudHQ and awesome partners. Details at https://t.co/skw7Rcit4c.
Stay tuned for:
🎉 Upcoming events
🚀 How to participate
😎 Win fantastic swag
Let's rock this Hacktoberfest! 💪🌈
#Hacktoberfest#opensource
Just wrote an narrative about the grey world and the non-grey world. The difference between them lies in whether we are forced to wait for a thing to show up.
A detective live in a non-grey world. Because their life's progress depend on whether they can find the target
This is a persistent thread about the progression of my new novel, it's written in Chinese but I will share the learning and funning thing in English under this thread.
#早餐一顆水煮蛋#AEggForBreakfast
The context here will be, the reason for adding some features is not to chase new customers but to make current customers happier.
This means if you don't have customers, this line is not for you
The essence of any successful product, as far as I can tell is simplicity. It always comes from saying no to something.
Hence, the death trap of start-ups is thinking "I will get more customers if I add this feature". It will form a death spiral of chasing endless features.
Has been busy implementing the new console of Instill-AI, I can't wait to show you our new design and workflow to build a pipeline to connect dozens of sources/destinations and models like OpenAI GPT-3.5 and Stability AI