zyklonix

Rosalia gives me hope... https://www.youtube.com/watch?v=htQBS2Ikz6c

I prefer this one: https://assets-prd.ignimgs.com/2023/10/06/colossus-the-forbi...

Recently discovered this movie. I still can't believe it was done in 1970! The visual props of the huge mainframes, tape drives, blinking lights, etc were real-computing gear from Control Data Corporation (CDC). Probably one of the greatest sci-fi movies of all time.

That’s a fair point, material abundance and consumer freedom have lifted billions out of hardship. What I found eye-opening in the essay, though, is that abundance can coexist with emptiness: when every decision is individual and transactional, we lose the shared meaning that once came from depending on one another. The argument isn’t against comfort or choice, but against mistaking options for connection.

I never really connected the modern idea of "freedom of choice" with today’s loneliness epidemic until reading this. The author makes a compelling case that our obsession with individual choice has quietly eroded the social fabric that makes freedom meaningful, turning independence into isolation.

That was fun!

I’m working on https://unrav.io

Building a new layer of hyper-personalization over the web. Instead of generating more content, it helps you reformat and interact with what already exists, turning any page, paper, or YouTube video into a summary, mind-map, podcast, infographic or chat.

The broader idea is to make the web adaptive to how each person thinks and learns.

Microsoft Songsmith is another one that deserved a second life. It let you hum or sing a melody and would auto-generate full backing tracks, guitar, bass, drums, chords, in any style you chose.

It looked a bit goofy in the promo videos, but under the hood it was doing real-time chord detection and accompaniment generation. Basically a prototype of what AI music tools like Suno, Udio, or Mubert are doing today, fifteen years too early.

If Microsoft had kept iterating on it with modern ML models, it could’ve become the "GarageBand for ideas that start as a hum."

I always thought Microsoft Popfly had huge potential and was way ahead of its time. It made building web mashups feel like playing with Lego blocks, drag, drop, connect APIs, and instantly see the result.

If something like that existed today, powered by modern APIs and AI, it could become the ultimate no-code creativity playground.

A high schooler writes about how absurd is a school system based on a fixed one-size-fits-all education and how AI is making this so obvious.

Totally agree, most prompts (especially for code) aren’t designed to surface novel insights, and even when they are, it’s hard to recognize them. That’s why the daydreaming loop is so compelling: it offloads both the prompting and the novelty detection to the system itself. Projects like https://github.com/DivergentAI/dreamGPT are early steps in that direction, generating weird idea combos autonomously and scoring them for divergence, without user prompting at all.

This idea of a “daydreaming loop” hits on a key LLM gap, the lack of background, self-driven insight. A pragmatic step in this direction is https://github.com/DivergentAI/dreamGPT , which explores divergent thinking by generating and scoring hallucinations. It shows how we might start pushing LLMs beyond prompt-response into continuous, creative cognition.

We might as well embrace them: https://github.com/DivergentAI/dreamGPT

Consider adding visuals on each node. Check out: https://explorer.globe.engineer/

Did you transform a coat stand into an awesome futuristic instrument??? So glad I asked. Brilliant work. This should be in the HN front page!

I am fascinated with this. What is it used for?

These are some great insights! Thanks for summarizing them. It will save me a lot of time in the future. I would love to see this digital ad! Is it public?

This reminds me of the famous "King - Man + Woman = Queen" embedding example. The fact that embeddings have semantic properties in them explains why simple linear functions would work as well.

Hallucinations are essential for divergent thinking. Not everything is solved following goal driven approaches. Check out DreamGPT: https://github.com/DivergentAI/dreamGPT

Summary: https://www.videogist.co/videos/jeff-dean-google-exciting-tr...

It only supports the Apple Magic Trackpad: https://support.apple.com/en-us/HT213998

That's great! It even supports VS Code. How did you manage to use the mouse? My AVP refuses to pair with a bluetooth mouse.

Any idea how the author was able to have multiple desktop screens? What app is he using for this? I thought it only allowed you to mirror one screen per Mac and from my experience it is quite laggy.

Thanks for the feedback. You do bring a good point there. I have updated the readme with a screenshot and more details on what is doing.

Thanks a lot! I highly recommend that you give it a try. You should be able to run it any PC/Mac. No GPU is required. It's fascinating the quality of the ideas that it generates. You can see a sample of what you get just on the first step ("dream" phase): https://github.com/DivergentAI/dreamGPT/blob/main/docs/img/o...

Yes, you can find a screenshot here (first "dream" iteration): https://github.com/DivergentAI/dreamGPT/blob/main/docs/img/o...

This is in fact the whole premise of Jeff Hawkins' book "On intelligence". Prediction is at the core of how we learn/act. Since we are born, we are just predicting everything that will happen and our brain learns when we see/hear/sense something that we can't predict.