self

152 stories and discussions about self, aggregated from every source we track.

1.

This site is now reachable over Tor as a hidden service, at a .onion address that resolves only inside the Tor network. Tor relays and encrypts your traffic as …

351 points•mooreds•about 2 months ago•116 comments•
2.
90 points•emurlin•about 2 months ago•31 comments•
3.

Large Language Models (LLMs) tend to add disclaimers like "I'm just an AI" when asked about something related to themselves. The self-reports from such responses are used in debates about AI safety or…

89 points•yu3zhou4•about 2 months ago•93 comments•
4.

September 12–14, 2027 in St. Louis, MO. A celebration of self-hosting and homelabs of all sizes.

64 points•geerlingguy•about 2 months ago•38 comments•
5.

There is nothing more satisfying than owning, end to end, the software and the hardware you use without relying on abusive cloud corporations. Internet is us, …

60 points•dalvrosa•2 months ago•40 comments
7.

The Harness of Harnesses • built for RSI: a trusted, persistent, self-evolving multi-agent ecosystem for all-domain collaboration. - EverMind-AI/Raven

53 points•cyfyifanchen•about 2 months ago•48 comments•
8.

Training the car to do self-parking using a genetic algorithm

46 points•trekhleb•about 2 months ago•5 comments•
9.

My literature search for recent work on composing self-stabilizing systems didn't yield anything useful. The layered stabilization idea was ...

43 points•matt_d•2 months ago•3 comments•
10.

A month ago, Tesla revealed several secrets regarding the new chip the Silicon Valley company has designed for full self-driving capability. Nonetheless, some of the people making that presentation may have failed to…

23 points•JAMorrow•over 7 years ago•32 comments
11.

AI scrapers inflated my visitor count tenfold. I now count a visit only after a user interaction, buffer hits on each web server, and back up the...

22 points•gioele•2 months ago•8 comments
12.

Healthchecks.io ping endpoints accept HTTP HEAD, GET, and POST request methods. When using HTTP POST, clients can include an arbitrary payload in the request body. Healthchecks.io stores the first 100kB of the request…

22 points•danlamanna•7 months ago•7 comments
14.

SVG viewer in SVG. Contribute to stevana/svg-viewer-in-svg development by creating an account on GitHub.

12 points•stevan•about 3 years ago•5 comments
15.

It’s been a while since we did a proper release of Self – too long in fact. The last duck themed release was “Mallard” a couple of years (!) back – we’re overdue…

11 points•doublec•over 9 years ago•0 comments
16.

A reference comparison of the self-hosted AI orchestrators in 2026: modalities, multi-machine support, auto-discovery, cache-aware routing, ops console, cloud burst, non-LLM fan-out, training, Kubernetes, platforms and…

10 points•nextime•2 months ago•0 comments•
17.

How to deploy a private, local AI gateway for engineering teams using LiteLLM, Caddy, Langfuse, and Open WebUI with multi-tier quotas.

9 points•marioezquerro•about 2 months ago•0 comments
20.
7 points•snippyhollow•about 2 months ago•1 comment•
21.
6 points•osr00•2 months ago•14 comments•
22.

Meta AI has released a new “Cookbook of Self-Supervised Learning," which provides AI researchers and practitioners with a practical guide on how to navigate SSL methods. This is part of our efforts to lower the…

6 points•mohanarpit•over 3 years ago•0 comments
23.

Parity’s engineers tested open-weight AI models through a self-managed inference stack. Over one month, 25 engineers generated nearly 13 billion tokens. Here’s what they learned.

5 points•ilovefood•2 months ago•0 comments•
24.

Sentience creates a unique model of your mind: an AI that remembers everything, thinks like you, and operates alongside you. It turns your email, meetings, messages, and screen into one private, encrypted memory you own.

5 points•tomaitagaki•2 months ago•4 comments•
25.

There is nothing more satisfying than owning, end to end, the software and the hardware you use without relying on abusive cloud corporations. Internet is us, …

5 points•dalvrosa•2 months ago•1 comment•
26.

The deployed fleet of ultra heavy-duty trucks and mining equipment assets powered by Volvo Autonomous Solutions reached a major milestone...

4 points•cisc•about 2 months ago•0 comments•
27.

Private offline AI on your phone. Run models locally and keep your chats on your device.

4 points•nathancheng•about 2 months ago•1 comment•
28.

Aldine: slim, self-hosted, open-source LaTeX collaboration (an Overleaf alternative). Real-time editing (Yjs CRDT), git-native branches, Zotero & GitHub sync, AI error fixing. Two containers, A...

4 points•rahloff•about 2 months ago•1 comment•
29.
4 points•geox•about 2 months ago•0 comments•
30.

There is nothing more satisfying than owning, end to end, the software and the hardware you use without relying on abusive cloud corporations. Internet is us, …

4 points•ibobev•2 months ago•0 comments•
31.
4 points•davletdz•2 months ago•0 comments•
32.

Contribute to mohitkumar188/AgentTrace development by creating an account on GitHub.

4 points•mohitkumar18•2 months ago•0 comments•
33.

Self-hosted cloud coding agents: Claude Code and Codex in per-thread sandboxes, one docker compose - dropalltables/valet

3 points•imbobbytables•about 2 months ago•1 comment•
34.

Create an AI workspace for your team with defined workflows, access controls, and human approvals. Work through AI chat and forms without technical expertise.

3 points•b4rtazz•about 2 months ago•0 comments•
35.

AI self-improvement may not be powerful enough to overcome diminishing returns. Even RSI may not lead to a fast takeoff to ASI (superintelligence).

3 points•gmays•about 2 months ago•0 comments•
36.

Your own workspace. A self-hosted, open source alternative to Google Workspace: mail, drive, docs, sheets, slides, calendar and chat. - eigen-is/eigen

3 points•0x0f0f0f•about 2 months ago•1 comment•
37.

A private AI appliance that talks, remembers, keeps your day, reads your documents and runs your home, on its own hardware. In Private mode (the default), nothing you say leaves the box.

3 points•phntmcore•about 2 months ago•0 comments•
38.

Contribute to google-research/rrsi development by creating an account on GitHub.

3 points•simonpure•about 2 months ago•0 comments•
39.

Open-source AI text detector — VirusTotal for AI slop. 23 engines, self-hosted, runs on CPU. Scan text or URLs locally. - pablocaeg/sloptotal

3 points•sloptotal•about 2 months ago•1 comment•
40.

Most developers still treat prompt injection as a leakage problem. Someone types an adversarial...

3 points•reidmarlow•about 2 months ago•1 comment
41.

If you follow AI safety, you know the uncomfortable truth: modern LLM guardrails are failing not...

3 points•kate8382•about 2 months ago•4 comments
42.

Teach it once, see what stuck, and watch verified memory improve a model served from a single iMac. cognira Entity runs it on your hardware. Free to use, no credit card.

3 points•thomasconway01•about 2 months ago•0 comments•
43.

July 2026 set a record for Tesla driver-assist crashes: 236, including 4 fatal ones the public never connected to Autopilot or FSD.

3 points•Markoff•2 months ago•0 comments•
44.
3 points•mathaix•2 months ago•3 comments•
45.

AI scrapers inflated my visitor count tenfold. I now count a visit only after a user interaction, buffer hits on each web server, and back up the...

3 points•zorrn•2 months ago•0 comments•
46.

Designed by myself – code built with AI – the code itself creates the music.One of my ongoing goals is to develop music-software-artifacts that educates the viewer on how the music itself is made –…

2 points•soundworlds•about 2 months ago•0 comments•
47.

AI self-improvement may not be powerful enough to overcome diminishing returns. Even RSI may not lead to a fast takeoff to ASI (superintelligence).

2 points•marojejian•about 2 months ago•1 comment•
48.

I'm already exhausted of hearing about Recursive Self Improvement

2 points•xondono•about 2 months ago•0 comments•
49.

SemIf (OpenJEV) semantic-if decisions served as a Hopsworks model deployment - MagicLex/jevworks

2 points•LexSiga•about 2 months ago•1 comment•
50.

Tech CEOs said they agreed on an AI safety “accord,” but it appeared voluntary, with Trump calling it “morally binding.” Trump also unveiled an AI-powered website, America.gov.

2 points•huijzer•about 2 months ago•1 comment•
51.

<p>What if your music server had a social network attached, and your social network could play music?</p>

2 points•blacklight•about 2 months ago•0 comments
52.
2 points•sc970•about 2 months ago•0 comments•
53.

Regularize the search, not the harness. Harness evolution that transfers out of distribution on a third fewer policy tokens.

2 points•jonbaer•about 2 months ago•0 comments•
54.

ASTRO — self-hosted, self-improving voice AI agent (PL): Raspberry Pi + Hailo NPU, wake word, STT/TTS, vision, LoRA self-training. Offline-first, open-core. - mobium-app/ai_astro_public

2 points•mobium-app•about 2 months ago•0 comments•
55.

Advances in language modeling have been driven by scaling pretraining on ever more data. Yet, the training data is still largely curated on the model's behalf. A more general approach to pretraining would let the model…

2 points•7777777phil•about 2 months ago•0 comments•
56.

Lean, self-hostable Calendly & Cal.com alternative — a single Go binary + SQLite, API-first and MCP-native. Apache-2.0. - Calnode/calnode

2 points•shockalotti•about 2 months ago•0 comments•
57.

A private AI appliance that talks, remembers, keeps your day, reads your documents and runs your home, on its own hardware. In Private mode (the default), nothing you say leaves the box.

2 points•phntmcore•about 2 months ago•2 comments•
58.

On-policy distillation (OPD) trains a student model by having it generate trajectories, then matching its next-token predictions with an external teacher's next-token predictions. This provides dense, token-level…

2 points•simonpure•about 2 months ago•0 comments•
59.

An essay on the self-correction mechanism and its civilizational consequence.

2 points•sebastianconcpt•about 2 months ago•0 comments•
60.

Advances in language modeling have been driven by scaling pretraining on ever more data. Yet, the training data is still largely curated on the model's behalf. A more general approach to pretraining would let the model…

2 points•E-Reverance•about 2 months ago•0 comments•

Related topics