Mo pfp

Mo

@meb

599 Following
595 Followers


Mo pfp
Mo
@meb
How I would define an AI agent at the start of 2025: In abstract, a piece of software that can achieve a certain goal, and uses AI to navigate the fuzzy path of achieving that goal. More specifically, a combination of - Goal - Context management - Underlying model - Tools (datastores and functions with effects) - Prompting & self-prompting strategies (including delegating to other agents) Example: A geocoding agent Goal: Parse content from a sentence and geocode it Context Management: Trimmed sliding window. We only need the very most recent messages to get the job done. Underlying model: A simple model like 4o mini is more than enough Tools: A function to search cities with inbuilt query match ranking, and a function with access to a Google Maps geocoding API Prompting: Single turn. The agent should be able to get the job done in a single exchange, or ask the user to clarify the location if needed
0 reply
0 recast
4 reactions

Mo pfp
Mo
@meb
Nice walk in Spandau Zitadel today Since weather is getting better again, who’s up for a Berliner farcasters walk / hike next Sunday ?
0 reply
0 recast
2 reactions

Mo pfp
Mo
@meb
Undervalued benefit of ChatGPT compared to humans conversations. You can always go back a step and not say the thing you just said
1 reply
0 recast
1 reaction

Mo pfp
Mo
@meb
Based meme
0 reply
1 recast
5 reactions

Mo pfp
Mo
@meb
Following up on yesterday's repo for learning AI agents, I just published a Substack article with a bit of walkthrough of what it's actually doing. This should provide better context. https://melboudi.substack.com/p/ai-agents-from-0-to-1
0 reply
3 recasts
3 reactions

Mo pfp
Mo
@meb
Tired of all the AI noise? I recently shipped an open source demo that takes you through the steps of agentic systems, from zero to hero. Feedback and PRs welcome! The goal of this repo is to educate technical and non technical people alike on the components of agentic systems https://github.com/motechFR/ai-agents-demo
4 replies
3 recasts
7 reactions

Mo pfp
Mo
@meb
Based dev tip of the day: Ask Cursor to fix a bug by deleting code instead of adding even more code
0 reply
0 recast
4 reactions

Mo pfp
Mo
@meb
Finally got my Apple Dev account approved this morning! Coming soon, my first iOS app, which of course has an AI running underneath. Will share the link.
1 reply
0 recast
7 reactions

Mo pfp
Mo
@meb
This is a real White House post. Memetic branch activation of the simulation confirmed https://x.com/whitehouse/status/1910072885789573505?s=46&t=heYf7m0kaq62UF_MFK0e6g
0 reply
0 recast
1 reaction

Mo pfp
Mo
@meb
Recent Google paper (February 2025) on prompt engineering. Super helpful https://www.kaggle.com/whitepaper-prompt-engineering
0 reply
0 recast
2 reactions

Mo pfp
Mo
@meb
Nearly done with v1 of my first iOS app using Cursor + AI, coming from zero native experience as a seasoned software engineer. What stands out: • UI/screens are fast to build. Cursor handles layout + SwiftUI boilerplate well. • Post-setup, bugfixing takes a few AI loops. Looking at the reasoning and debug tokens is uncanny valley territory. • Asking the AI “why” helps me level up, but I still need docs + web deep dives. • I now reach for docs more than before, faster than waiting on the AI sometimes. • Prior software/product experience helps frame prompts + architecture. • Server-side in a language I know = security peace of mind. TL;DR: • Pro engineers: still extremely relevant. AI boosts productivity, especially in custom, less-trodden paths. • Copy/paste devs: rugged
0 reply
0 recast
1 reaction

Mo pfp
Mo
@meb
Postgres really can do everything meme lives on. Working on an app that involves generating audio. Nearly opted for classic upload / download with S3. Then I remembered my average filesize is 50kb. Enter postgres. I was able to create a table with a bytes column, that works really well. PSA to keep your infra as simple as needed, and not reach out for more difficult solutions till necessary.
4 replies
0 recast
14 reactions

Mo pfp
Mo
@meb
Until recently, gpt 4o mini has been my go to for lots of agents given its low cost and latency. Did some tests today with Gemini 2.0 flash, which somehow is even less expensive, but supposed to be full 4o parity and got amazing results, especially with it following instructions very well. Shoutout to @alexpaden for sharing the tip on xml-ish tags for your prompts. Seems to be more machine readable, but also makes it way easier to structure
1 reply
0 recast
2 reactions

Mo pfp
Mo
@meb
Today, for the first time in a while, I solved an issue by going from stackoverflow, to Reddit, to the official docs. Nostalgia vibes
1 reply
0 recast
3 reactions

Mo pfp
Mo
@meb
First time vibe coding yesterday (swift in cursor). Very cool experience and moved ahead incredibly fast compared to the old days of reading the docs, and moving up the stack from "hello world" when working with new languages. I don't think all devs are cooked yet, but this has already passed the level of the average "copy paste stack overflow without caring" dev. What I'm enjoying: 1. Being able to spend more time in product land, versus figuring out syntax. 2. Reducing the barrier to porting my engineering skills to a different language. I think LLMs will be an awesome supporter for the mobility of experienced engineering talent between language ecosystems. 3. Setting up boilerplate and screens way faster, so I can focus on iterations Speedbumps: 1. Architecture. Getting the AI to follow rules such as decomposing screens into different files, and explain certain design decisions it makes. 2. Verbosity. Feels like too much written some times. I generally have to prompt for simplicity.
0 reply
1 recast
1 reaction

Mo pfp
Mo
@meb
Looking to get into iOS app development Go for swift or react native? Not intending to publish to android (for now)
7 replies
1 recast
4 reactions

Mo pfp
Mo
@meb
Saved some time using AI to translate an English doc to French. The canonical google French version for Farcaster is "Farcasteur"
2 replies
0 recast
5 reactions

Mo pfp
Mo
@meb
Feature request inspired by situation I just had; Can we get LinkedIn post embeds like we have with Twitter? Pros: - Enable people targeting clients outside farcaster world to reshare content here - Extend the farcaster social graph back into places like LinkedIn - Build the bridge between normies and crypto natives
2 replies
0 recast
2 reactions

Mo pfp
Mo
@meb
AI Knowledge Graph productivity hack Send me some <3 on LinkedIn https://www.linkedin.com/feed/update/urn:li:activity:7311890866313297920/
0 reply
0 recast
2 reactions

Mo pfp
Mo
@meb
Following on yesterdays chat about Knowledge Graphs I discovered Microsoft recently released a framework that reverse engineers knowledge graphs from unstructured text https://youtu.be/6vG_amAshTk?si=iG31IpbvsCZZ9dx4
1 reply
0 recast
10 reactions