Tristan Rhodes (@tristanbob) 's Twitter Profile
Tristan Rhodes

@tristanbob

Vibe coder, AI engineer, open source advocate, network engineer, game dev, trumpeter, and dad. Building Cherishable.ai

ID: 71355794

linkhttp://cherishable.ai calendar_today03-09-2009 20:51:00

6,6K Tweet

2,2K Followers

5,5K Following

Tristan Rhodes (@tristanbob) 's Twitter Profile Photo

I love the incredible features and low-cost backend services provided by Supabase. That said, the FREE plan limit of two active projects can be frustrating at times. (BTW, I also have a paid project) I create new experiments every week. This lets me quickly test ideas, learn

Tristan Rhodes (@tristanbob) 's Twitter Profile Photo

I already have a great job that pays the bills. Most of the work you see on this account is driven by my passions. I have a passion to learn new skills, to build things, and a fascination with trying to understand how AI will change society. It's also largely driven by the

Tristan Rhodes (@tristanbob) 's Twitter Profile Photo

.Appwrite is an opensource backend provider similar to Supabase, but with a different stack that includes MariaDB and redis. What is your experience with Appwrite?

Tristan Rhodes (@tristanbob) 's Twitter Profile Photo

I love this description of how fast things have changed due to AI, and how different the people entering the workforce will think and operate.

Tristan Rhodes (@tristanbob) 's Twitter Profile Photo

There is a new pattern developing in vibe coding industry. This pattern involves letting users change the "mode" of the AI agent. Modes change the system prompts and control what resources and tools the AI agent can use. I think this trend started with Lovable when they

There is a new pattern developing in vibe coding industry.

This pattern involves letting users change the "mode" of the AI agent.  Modes change the system prompts and control what resources and tools the AI agent can use.

I think this trend started with <a href="/lovable_dev/">Lovable</a> when they
Tristan Rhodes (@tristanbob) 's Twitter Profile Photo

I’m having problems with my newest vibe coding project. This is another backend service that integrates several existing cloud services. For the very first time, I’m trying to perform testing immediately after implementing each feature. I’ve learned about unit testing,

Tristan Rhodes (@tristanbob) 's Twitter Profile Photo

I tested Jules (Jules) today. Jules is Google's AI coding agent that delivers on the vision that was first introduced to the world by Devin. To be honest, I'm still not sure what to make of it. It is the most autonomous coding tool I've ever used. Similar to vibe

Tristan Rhodes (@tristanbob) 's Twitter Profile Photo

I already loved NotebookLM for their audio overviews, but soon you will be able to create video overviews! There are so many interesting use cases for this! Teachers, wouldn't it be nice if you could turn your lesson plans into a short video for your students to watch?

Tristan Rhodes (@tristanbob) 's Twitter Profile Photo

You know things are getting serious when posts like this are not an April Fool's joke. Yes, DolphinGemma is real and tries to understand the sounds dolphins make.

Tristan Rhodes (@tristanbob) 's Twitter Profile Photo

I’m very curious about this trend of releasing AI coding models. Windsurf, OpenAI, and now v0. Will specialized AI models make a comeback? We all thought so at first, but none of them could keep up with the improvements in foundation models.

Tristan Rhodes (@tristanbob) 's Twitter Profile Photo

All of these videos were made with Google AI Veo 3. (Sound on!) In addition to the best video generation we have seen so far, all of the audio (sound effects, dialogue, and music) was automatically generated by Veo.

Tristan Rhodes (@tristanbob) 's Twitter Profile Photo

We should all be cheering and celebrating the companies who pay developers to make open source software. Let’s reverse the mindset that money in open source is bad!

Tristan Rhodes (@tristanbob) 's Twitter Profile Photo

I'm testing @Upstash Vector, a hybrid semantic search (RAG) with both sparse and dense vectors. I love that Upstash provides sample datasets to get started quickly. This dataset is 793 records and 1.7 MB (without the vectors). I generated this test using Upstash's helpful Data

I'm testing @Upstash Vector, a hybrid semantic search (RAG) with both sparse and dense vectors.

I love that Upstash provides sample datasets to get started quickly. This dataset is 793 records and 1.7 MB (without the vectors).

I generated this test using Upstash's helpful Data