The final missing piece for AI Assistant advancement (including Agents) is memory. Feels like we’ve improved a lot on reasoning, speed, quality etc… but the one final piece of the puzzle that I personally haven’t seen much on is memory. How do we allow a long term memory to

I don’t think there is a single piece of content more informative on American Foreign policy like this is. If you have time it’s worth the watch and understanding youtu.be/ava7rqirOYI?si…

Walked through a castle today and thought, ‘This would absolutely slap on horseback with chainmail and a broadsword’

Step by step making Otto affordable for everyone 🫡

I’ve always played around with the cursor agent but was never a huge fan of it tbh, just because I felt like it ruined my flow. Like I’d be mid coding and it would sit there for a while doing work and kills my focus. But now im starting to see where it can be useful. Took me a

how the fuck do I price an agent for someone putting 100 files of 10000 pages each vs. 1000 files of single pages?? Or trying to do something with a 50k token prompt vs doing 1M things with 1k token prompts ??? I think with where we’re going with AI, it’s gonna be hard for it to

Damn Claude 3.7 just ain’t it for coding I guess…

What if agentic memory was just a database with agentic querying ?…🤔 Hear me out, what if long term memory is just infinitely scalable table(or tables), but just queried using an agent that can continue to query intelligently and refine the query in real time until you get

Decided to just start playing with agents and knowledge graphs, onto a pretty good start 🤔

Recently hopped on substack, anyone with good content worth subscribing to? Can be about anything, not just AI

my chatgpt “memory” has been at 100% for like a whole year now. I found that super useful and was annoyed I couldn’t just get more. So got bored and decided to work on fixing this myself and building an agentic long term memory. I think I actually finally got it to work. I ended

Is it just me or does chat gpt have a bit of an attitude or a slight bit of a personality when you talk to it? Not full blown out personality but almost like hints of it ?

This is what I mean when I say ChatGPT has gotten so good at talking and is starting to feel a lot more personal. It’s a bit cringy/cheesy at times but it feels like it’s starting to have a personality

Jesus Christ, top roasts of 2025 so far

We’ve reached extremely dangerous levels of capitalism aka late stage capitalism. Levels where the government is purely just an oligarchy. Where laws and bills being passed aren’t for the people, they’re for the oligarchs who run it. Anything being pushed, voted on, or even

I think knowledge graphs in agentic work is so undervalued and utilized atm…

Really stoked to see Amjad Masad on the Joe Rogan podcast. Super grateful to have had him as an early investor at Cognosys. His story is inspiring and it makes me proud to see fellow Arab and especially Jordanians pushing tech forward and putting our part of the world on the map.

Sorry but is GPT 5 just ass ? I’m not even asking anything complicated. Just trying to have a a regular simple conversation. It almost feels like OpenAI tried to move the model from being “assistant” like to more human like but didn’t even do it well. Not too sure how I feel