Feed aggregator

Ten years ago I built a small evolutionary toy experiment with two types of agents: selfish and cooperative “ducks”.

At first, selfish strategies dominated. But when agents were given memory — the ability to remember who helped them — cooperation suddenly became stable under resource scarcity.

That experiment stayed in the back of my mind for years.

Recently I started rebuilding the idea from scratch as a larger system:

BiomeSyn

https://biomesyn.com/

Instead of evaluating AI on static tasks, the goal is to explore long-horizon adaptive environments where agents must:

• gather resources • survive environmental pressure • compete with other agents • adapt over many generations

The system is deterministic, so experiments can be reproduced across seeds — which makes it possible to treat it as a benchmark for adaptive agents.

The bigger question I’m interested in:

> What happens when intelligence is evaluated inside a world that keeps evolving?

Many current benchmarks measure short-episode performance. But real adaptive systems must operate in open-ended environments.

BiomeSyn is still an early research sandbox, but I’m curious whether environments like this could become useful for studying:

• evolutionary computation • long-horizon RL agents • multi-agent ecosystems • adaptive AI systems

Would be interested to hear thoughts from people working on agents, simulation platforms, or large-scale AI systems.

https://biomesyn.com/

Comments URL: https://news.ycombinator.com/item?id=47270182

Points: 1

# Comments: 0

Built a PWA where anyone can drop a pin on a live shared map to report nearby incidents — crime, fire, disasters, civil unrest, infrastructure failures.

No accounts, no tracking. IPs are one-way HMAC-hashed before storage and never logged raw. Community votes incidents from pending → active so noise gets filtered out automatically. Each category has its own TTL so the map stays fresh.

Stack: Next.js 16, Firebase (Firestore + Cloud Functions + FCM), Leaflet for the map, D3.js for a 3D globe view. Fully open source.

GitHub: https://github.com/BaselAshraf81/threatalert

Comments URL: https://news.ycombinator.com/item?id=47270181

Points: 1

# Comments: 0

Article URL: https://btw.media/

Comments URL: https://news.ycombinator.com/item?id=47270179

Points: 1

# Comments: 0

Whiplash sits in your macOS menu bar and gives you a live overview of all your concurrent agent sessions and their status across iTerm, Ghostty, or Terminal. A little open source, utility that I hope helps you too.

Comments URL: https://news.ycombinator.com/item?id=47270152

Points: 2

# Comments: 0

Article URL: https://reorchestrate.com/posts/your-binary-is-no-longer-safe-conversion/

Comments URL: https://news.ycombinator.com/item?id=47270149

Points: 1

# Comments: 0

Article URL: https://reorchestrate.com/posts/your-binary-is-no-longer-safe-decompilation/

Comments URL: https://news.ycombinator.com/item?id=47270146

Points: 1

# Comments: 0

Article URL: https://www.theregister.com/2026/03/06/ai_kills_software_licensing/

Comments URL: https://news.ycombinator.com/item?id=47270122

Points: 8

# Comments: 0

Article URL: https://theintercept.com/2026/03/05/kosa-online-age-verification-free-speech-privacy/

Comments URL: https://news.ycombinator.com/item?id=47270104

Points: 5

# Comments: 2

Article URL: https://www.youtube.com/watch?v=O0pFqhZl3fg

Comments URL: https://news.ycombinator.com/item?id=47270088

Points: 1

# Comments: 0

Article URL: https://github.com/TinyAGI/fractals

Comments URL: https://news.ycombinator.com/item?id=47269860

Points: 1

# Comments: 0

Article URL: https://clickhouse.com/blog/clickstack-embedded-clickhouse

Comments URL: https://news.ycombinator.com/item?id=47269854

Points: 1

# Comments: 0

Article URL: https://www.amazon.com/dp/B0GR6T1KMQ

Comments URL: https://news.ycombinator.com/item?id=47269831

Points: 1

# Comments: 0

Article URL: https://mapscaping.com/as-the-crow-flies-distance-calculator/

Comments URL: https://news.ycombinator.com/item?id=47269810

Points: 1

# Comments: 0

I keep hearing the same thing across aerospace/defense and other regulated B2B programs. Even when both companies use Jira internally, the moment a customer (or other external partner) suggests “let’s just share a Jira project,” it turns into a weeks/months-long IT + infosec ordeal… so teams fall back to email + Excel trackers.

If you’ve lived this, I’d love detailed stories. Some conversation starters:

>> What exactly made it hard? (SSO/IdP, user provisioning, domain policies, MFA, VPN, IP allowlists, Atlassian Access, SCIM, contractors, etc.)

>> Is the blocker usually IT, security, compliance, procurement/vendor risk, or the Jira admins themselves?

>> Jira Cloud vs Jira Data Center: which is worse for external collaboration and why?

>> What are the common “policy red lines” that cause a hard no? (least privilege, separation of tenants, auditability, data residency, CUI/ITAR, SOC2, etc.)

>> What workarounds did you end up using instead (shared spreadsheet, shared mailbox, separate “shadow Jira,” Confluence page, etc.) and what broke?

>> If you did make cross-org Jira work, what was the setup that finally passed and how long did it take? If you didn't make it work, what happened?

Context: I’m trying to understand the true root causes and failure modes -- whether this is mostly technical (identity + permissions) or mostly organizational/policy, and what parts are actually solvable.

Comments URL: https://news.ycombinator.com/item?id=47269793

Points: 1

# Comments: 0

Article URL: https://reason.com/2026/03/04/computer-scientists-caution-against-internet-age-verification-mandates/

Comments URL: https://news.ycombinator.com/item?id=47269782

Points: 1

# Comments: 0

I built SlideScholar to solve a pain point I kept seeing: researchers spending 6-10 hours making slides for 15-minute conference talks.

Upload a PDF or ArXiv URL, pick your talk length, and get an editable .pptx presentation with assertion-evidence titles, extracted figures and tables, and speaker notes with timing cues.

Stack: Next.js frontend on Vercel, Python/FastAPI backend on Railway, Claude API for content planning, python-pptx for slide generation.

Free, no signup needed

Comments URL: https://news.ycombinator.com/item?id=47269745

Points: 1

# Comments: 0

I rarely have a customer experience that genuinely feels delightful. Lately I've started to wonder why that is... I get up sold on products when I'm not ready to buy, I receive emails about features I'd never use, and when I want have an issue it's impossible to find someone to talk to. This dynamic always struck me as weird.

But having worked for large brands, the truth is that most companies have no idea when and how to talk to their customers. They rely on a messy web of conflicting events and triggers that engage customers without context.

I’ve recently started working on that tries to fix this, by pulling customer events from every channel, derive important moments from sequences of events and trigger the right engagement based on your specific context, all while continually learnig what's important to customer and how to best engage them

Here's how this works in an e-commerce example (although this works for any type of brand): 2 customers may have abandoned their cart at checkout. Customer1 got an error during checkout, got frustrated and moved on, Customer2 just a regular session.

Today companies treat these 2 customers the same and just send a discount code after 24h when in reality you should investigate customer1's issue and let them know it was fixed so they can complete their transaction

I call these sequences of events that drive customers to do X vs Y, "moments". My thesis is that you can discover these moments and design customer engagement around them to build delightful experiences that feel like you're going above and beyond, tailored to the customer which improves revenue, retention and advocacy.

I would love to hear from anyone that has experience with this problem.

You can follow my journey at https://booly.co

Comments URL: https://news.ycombinator.com/item?id=47269729

Points: 1

# Comments: 0

Article URL: https://github.com/openai/symphony

Comments URL: https://news.ycombinator.com/item?id=47269722

Points: 1

# Comments: 0

"Remote Control" lets you watch your session from the couch. commuter lets you take it to work.

Comments URL: https://news.ycombinator.com/item?id=47269704

Points: 2

# Comments: 0

Article URL: https://octopress.org/2015/01/15/octopress-3.0-is-coming/

Comments URL: https://news.ycombinator.com/item?id=47269699

Points: 1

# Comments: 2