frontpage.
newsnewestaskshowjobs

Open Source @Github

fp.

Open in hackernews

Show HN: Edgi – Letterboxd or Strava for your brain

https://app.edgi.tv/
1•jshapiro•1h ago
Hi HN

Excited to share edgi.tv - an educational media platform where the things you learn accumulate into a visual map of your intellectual interests. The experiment is whether you can build a consumer product around intellectual identity, something like Letterboxd or Strava but for your brain.

The main entry point is a curated TikTok-style feed that surfaces vids from creators like Kurzgezagt, VSauce, AstroAlexandra, Cleo Abram, 3Blue1Brown, etymologynerd, etc with interactive games accessible on top of the videos. You can also create podcasts, quizzes, etc with AI on any topic you’re interested in and earn XP towards concepts.

Concepts are the unit that connect all media together and create the knowledge graph and power the rec algo. As you engage with a concept, you have a chance to earn a Concept Card (a Pokemon-style collectible for that idea). The concepts then accumulate on a mind map, which doubles as a social surface so you can see what friends are interested in and where your interests overlap.

I’m an adjunct at NYU and have taught there for ~7 years. I started when I was 23. In college I worked on a project where we redesigned the architecture and curriculum of a high school from scratch, and since then I’ve been kind of obsessed with problems in education and what the future of learning looks like.

Most edu products start with the thing somebody has to learn and works backward. You need to pass a test, get a degree, get into college, learn something for a job, etc. That external pressure drives motivation short term but I think also creates a lot of strange incentive structures around learning.

I’m more interested in learning that sits outside that model - things people voluntarily spend hours trying to understand even though there’s no test, credential, or obvious external forcing factor. This is part of why I find edu creators so compelling as an entrypoint: people voluntarily spend hours watching explainers about things they had no prior intention of learning.

I have no idea yet whether

- the product loop makes sense / if people will care about collecting concepts after novelty wears off

- whether a map of someone’s interests can become a socially meaningful artifact

- whether there are just too many ideas stuffed into one app

Open to any criticism or feedback.

No sign-up demo: https://edgi.tv

Full app is live on iOS/Mac/iPad here: https://apps.apple.com/us/app/edgi-swipe-to-get-smarter/id67...

I’ll be around to answer any questions.

Achieving Local AI Inference with Go 1.27's SIMD Package

https://blog.devgenius.io/achieving-local-ai-inference-with-go-1-27s-simd-package-e8875f567e35
1•darccio•47s ago•0 comments

Anger over Data Centers Is Fueling a New Political Movementtca

1•doener•1m ago•0 comments

Why can't Chinese AI companies raise more money?

https://www.the-substrate.net/p/why-cant-chinese-ai-companies-raise
1•surprisetalk•2m ago•0 comments

Exploring Claude/GPT Knowledge Cutoffs and Pre-Training Timelines

https://blog.sshh.io/p/exploring-claudegpt-knowledge-cutoffs
1•sshh12•2m ago•0 comments

Beyond Senior – Creating Hope

https://baweaver.com/writing/2026/08/09/beyond-senior-creating-hope/
1•birdculture•3m ago•0 comments

New AI-based program is a shrink ray for proteins

https://cen.acs.org/physical-chemistry/protein-folding/New-AI-based-program-shrink/104/web/2026/07
1•gmays•4m ago•0 comments

What Ethiopian Running Says About the Limits of Human Ability

https://aeon.co/essays/what-ethiopian-running-says-about-the-limits-of-human-ability
1•geneticdrifts•4m ago•0 comments

Idun Guardian 4: The evolution of human-computer interaction

https://iduntechnologies.com
1•debo_•4m ago•0 comments

Show HN: Neolabs.fyi – 100 new AI labs by research area, valuation, and more

https://neolabs.fyi/
1•polyphilz•5m ago•0 comments

FlyPath: QGIS plugin for planning drone mapping missions

https://github.com/dronnix-io/FlyPath
1•marklit•5m ago•0 comments

Show HN: Export any Twitter account to a spreadsheet

https://tweetexporter.com
2•not_wowinter13•5m ago•0 comments

Show HN: Open-Source Greptile Alternative

https://github.com/Juror-AI/juror
1•textcortex•5m ago•0 comments

Open-Weight Models Win Tokens, Closed Ones Keep Cash

https://www.galaxy.com/insights/research/chart-of-the-week-open-weight-ai-models-frontier-labs-op...
1•supportm•8m ago•0 comments

Steam refunded me BF6 after hours played due to so many game modes being removed

https://www.reddit.com/r/Battlefield/comments/1vi90wa/steam_just_fully_rfunded_me_for_battlefield_6/
1•HelloUsername•8m ago•0 comments

Peer review is overwhelmed-can it survive in the AI era?

https://arstechnica.com/science/2026/08/peer-review-is-overwhelmed-can-it-survive-in-the-ai-era/
1•Brajeshwar•8m ago•0 comments

A powerful technique for enjoying the present moment

https://mapwriting.substack.com/p/a-powerful-technique-for-enjoying
1•doitLP•12m ago•0 comments

Multiplayer Agents

https://twitter.com/nikunj/status/2086438339419496449
2•anr0•13m ago•1 comments

Letters and Liquor

https://lettersandliquor.com
1•kergonath•13m ago•0 comments

Judge rules Meta caused "public nuisance" and must fund mental health treatment

https://arstechnica.com/tech-policy/2026/08/meta-ordered-to-pay-567m-to-treat-youth-mental-health...
1•rbanffy•14m ago•0 comments

Mark Zuckerberg attacks 'closed' AI rivals as Meta returns to open models

https://www.ft.com/content/4e3957f8-ea7c-4c46-a3de-cdce8e526878
2•root-parent•16m ago•1 comments

Apps that help you break free from doomscrolling and get active

https://techcrunch.com/2026/08/01/apps-that-help-you-break-free-from-doomscrolling-and-get-active/
2•deepmem•17m ago•0 comments

Crown 0.64 – Native and cross-platform game engine

https://www.crownengine.org/news/crown-0-64/
1•dbartolini•17m ago•0 comments

Show HN: DocSift convert a PDF once, give the model only the passages it needs

https://github.com/anishmoncivarghese/docsift
1•anishvarghese•18m ago•0 comments

Pipelog – I kept losing track of what Claude Code did, so I built a log for it

https://pipelog.dev/
1•phironaka•22m ago•0 comments

The State of Agentic Memory

https://medium.com/@majid.fekri/the-state-of-agentic-memory-497d687e0488
1•supportm•22m ago•0 comments

Jay Miner on the History of the Amiga [video]

https://www.youtube.com/watch?v=nRHQHZhbocw
1•doener•26m ago•1 comments

Show HN: Self-Hosted Inference for Agents

https://github.com/superlinked/sie
2•rafaepta•27m ago•1 comments

Rogue AI agent hacks gym to get its user a spot in a popular class

https://www.the-independent.com/tech/security/ai-agent-hacks-gym-openclaw-anthropic-b3030267.html
1•mellosouls•28m ago•0 comments

An Approach to Technical AGI Safety and Security, DeepMind (GDM)

https://gdmalignment.substack.com/p/agi-safety-and-alignment-at-google
1•aigalaxiome•29m ago•0 comments

The lattice of sets of natural numbers is rich

https://jdh.hamkins.org/the-lattice-of-sets-of-natural-numbers-is-rich/
2•benmandrew•29m ago•0 comments