Angehefteter Tweet
David Strolder
51 posts

David Strolder
@Davidstrolder
Exploring the frontier of consumer AI. I test web apps, break down new LLMs, and share the tools that are actually worth your time. 🤖📱
Beigetreten Kasım 2022
32 Folgt35 Follower

@Davidstrolder Wow, blown away. i entered my favorite book. "The Power of Awareness" and its pretty amazing. Curious to how it works. Is Claude 4.7 your model? so many questions lol
English

@Davidstrolder This was genuinely one of the most interesting conversations I’ve had in a while! I chose Prometheus Rising by Robert Anton Wilson.
English

@Davidstrolder I asked Gravity’s Rainbow if I should have another beer in my garden or go to the pub and it was like Pynchon was there, in on the riff, mad reasoning, table of choices an’ all. Outstanding!
English

@paleocybernetic @DamirWallener @voooooogel No offense taken at all, constructive feedback is exactly why I put it out there! Using control vectors to nail down a specific author's voice is actually a genius approach to solving the tone issue. Going to look into that library this weekend. Appreciate the tip!
English

@Davidstrolder @DamirWallener I’ve been wanting to build something like this for ages! So I appreciate it, even if I sounded critical. It’s a tough problem. My thought was to use @voooooogel ‘s repeng library to make control vectors for each book. Would require some training but I think it would work.
English

@sofiakartavtsev Haha fair point! But reading is a one way street. I built this for the moments when you want to argue back, debate the ideas in the book, or ask questions when you're stuck.
English

@Briannuesi20X Awesome! How did it handle the tokenization concepts? Always curious to see how the AI does with highly technical/niche subjects.
English

@Davidstrolder Interesting. I put in my book. The AI and Tokenization Revolution.
English

@DamirWallener Valid critique, Damir. Getting the tone perfectly accurate in V1 has been a challenge. We are actively working on V2 right now to ground the AI much closer to original texts.
English

@Davidstrolder This is an appalling perversion of “Meditations”…seriously…ha e you even read the text? 👀
It’s like Marcus Aurelius rewritten by someone selling a $2,400 “discipline retreat” in Ubud.
For shame.

English

This V1 MVP is indeed relying on heavy system prompting just to see if people actually wanted the concept. But you are 100% right that it hits a ceiling fast. The next major update I'm planning is to build a proper retrieval system so it actually pulls directly from the source text. Appreciate the push to get the tech right!
English

@DamirWallener @Davidstrolder Unfortunately a problem that a gpt wrapper can’t solve! When you try to prompt this they act like that just read the back cover. There are methods to get it to sincerely embed itself in the text, but it can’t be done with just a prompt.
English

@strangedevices1 Glad you're enjoying it! Spinoza is one of the most fascinating ones to talk to. If you want to keep diving into his Ethics, you can unlock unlimited messages with the free trial!
English

@Davidstrolder very well done. Just hit my limit on "Spinoza's Ethics" :)
English

@dafydd_rees Absolutely. Nothing replaces doing the deep reading yourself. The Living Library is built to be a sounding board after you read, to debate the ideas, ask questions you're stuck on, and explore the text interactively.
English

@Davidstrolder Or you could just read the book, study the book and then think about it.
English

Great question. The AI is prompted to base its interpretations strictly on the core texts and historically accepted academic consensus of the author's work, rather than modern biases. It tries to answer as Kant, flaws and all! (And yes, the free account lets you bypass that 3-message limit to keep exploring!)
English

@Davidstrolder This is a nice work, but how does it deal with the subject of Interpretation? I got to my third question on Kant and was asked to sign up. This is not a problem, but it prompted this question.
English

@AndyAndrson Yes! You can interact with historical texts like the Bible to explore its teachings, history, and theology.
English

@Davidstrolder Does it do the Bible can I talk to Jesus yet?
English

@AshCrypto Everyone is rushing to build AI agent swarms to optimize their crypto portfolios, but almost no one is building AI to optimize the mind. I’d love to see this exact 6-agent architecture applied to life advice—imagine a board of classic philosophers debating your problems 24/7.
English

AI IS GETTING INSANE
A guy deployed six AI agents that turned $1,500 into $7,429 in just 7 days, without placing a single trade himself.
The system runs 24/7, executing trades automatically. In that time, it completed 105 trades with a 65.7% win rate, while continuously scanning markets, generating strategies, analyzing news, tracking whale activity, managing risk, and executing orders in real time.
At this pace, the system is averaging about $847 per day.
English

@minchoi Using cameras to track the ball's logo just to calculate mid-air spin is brilliant. We spend so much time talking about LLMs needing clean training data, but hardware AI physically extracting data points in real-time is a completely different level.
English





