22 year old, AI Master at Utrecht University Founder https://xitter.cf/t.co/P6Jys9syNZ - 550 MRR Co-founder https://xitter.cf/t.co/73fdUj3ZUp - 0 MRR AI enthousiast & advocate
Netherlands
Joined October 2024
- Tweets2K
- Following1.5K
- Followers1.3K
- Likes1.2K
Pinned Tweet
Which advanced AI feature would you like to integrate into your system most?
Drop your ideas below 👇 .
I will pick three of them and build them for free with some free credits for you.
Really looking for feedback for my API infrastructure.
Will actually put in effort!!
1M tokens for what?
Opus 5.5 is great, but it does feel very slow in a way and spending lots of time and tokens on things that may not be neccesary.
What is the best way in which LLMs can improve?
My opinion: Entity embeddings able to understand not just classes but individual entities, such as companies and people in much more depth.
My girlfriend wanted to publish a book she has been writing for 4 years.
Still lots of challenges in publishing books so I created an AI powered publishing company with her.
Full per character AI voiced audiobooks, with music and more.
Buy her book!
enqualiabooks.com/crazydaisy
Asked Claude Fable to get my website launch ready and it got flagged when checking security, interesting to see!
Maybe have to start using X again as I can't use Claude Code, what is going on here, what are you guys working on?
Still feels like Opus 5 is signficantly worse than Fable 5, using Opus 5 for most tasks now, but it just kept failing and failing, Fable 5 sovled it in minutes
Pretty intersting to ask Claude Code to construct your family tree. It's truly very impressive what it can find!
I see a lot of negativity surrounding Claude Code on my timeline, which kind of suprises me.
The product is amazing, and 50% Fable limits are not as bad as it sounds, as it should mainly figure as a director sending of Sonnet subagents which works perfectly.
Beginning July 20, Claude Fable 5 will be included in all Max and Team Premium plans, at 50% of limits.
Pro and Team Standard users will continue to have access to Fable via usage credits, and will receive a one-time $100 credit.
Demand for Fable has been challenging to predict, which is why we rolled it out to subscription plans in stages, extending access several times as we secured additional capacity.
Brainstorming for my Masters thesis:
What concepts do LLMs understand poorly which would improve the way AI can be useful?
Example: LLMs do not naturally understand employees well, what their strenghts are for instance.
Finally upgraded to 20x Max.
Time to lock in for summer now I'm free from University!
Excited to share the journey! Mainly focussing on growing my custom API SaaS enhanced by my own GraphRAG implementation now!
I could actually see this happening.
We will have agents that learn your type.
They will argue with agents of others if you are a compatible match and setup a date using your schedule.
You just get a notification: “You have a date at this coffeeshop on Friday”
Would you?
Starting to think about my Master thesis in AI.
Any recommendations?
Thinking about a GraphRAG system enhanced by embeddings