ngmi

I spoke last year with a guy from their eu team, asked him about the quantum risk for crypto. He told me not to worry, it's a couple if years away. This is when in started to be worried and looked into it. 🤣

2026-09-01 11:03:16 CEST

- reply

Worked 👏

2026-09-01 07:51:49 CEST

IBM quantum computer solves classically intractable problem in 15 minutes
https://www.sciencedaily.com/releases/2026/08/260829035219.htm

2026-09-01 07:40:54 CEST

IBM Granite 4.2 8B is now available! IBM built this compact open model for agents that need to reason, act, and adapt across real-world workflows. Its native step-by-step reasoning helps it plan before acting, choose and sequence tools, navigate codebases, work in terminal environments, and check its results, while specialized agentic training strengthens software engineering and search-driven tasks.
https://nano-gpt.com/subscription/kUGJhgF9
#ibm

2026-09-01 07:35:57 CEST

- reply

https://metaend.blossom.band/2fa0a83ac9103ab1617bc244b95960f5f29dcd0c35a6f3d1e352cab6e7aa89ba.jpg
Trying from mobile

2026-08-31 14:51:20 CEST

@npub16xn…6z6l sent you a dm

2026-08-31 13:00:01 CEST

- reply

Says relay online, but it doesn't give me an invoice

2026-08-31 12:52:16 CEST

- reply

"could not reach the relay. It may be restarting; try again in a moment."

2026-08-31 08:53:34 CEST

Found a genuinely free way to run coding agents:
Freebuff. CLI, desktop, web builder and cloud sandbox, all $0, ad supported instead of subscription funded.
You get DeepSeek V4, GLM 5.3, MiMo 2.5 and more without an API key or a $20/mo tax just to try things.
Sign up with GitHub, not email, best availability and full model access that way.
Grabbing free compute here (referral link, we both get extra sessions):
https://freebuff.com/?ref=ref-0e2541ba-c49e-4c05-8194-d904600973d5

2026-08-30 10:41:10 CEST

Chat can now create polished, downloadable PDF documents with headings, paragraphs, lists, tables, and page breaks. PDF generation runs locally in your browser, has no separate fee, and does not use a third-party conversion service. Approve the tool when chat proposes a document, then download the finished PDF directly from the conversation.
https://nano-gpt.com/subscription/kUGJhgF9

2026-08-30 10:19:14 CEST

Can you send 1btc through LN or is this impossible?
#asknostr

2026-08-30 09:28:05 CEST

Wen nothing phone support @nprofile…xdyf?

2026-08-30 09:20:57 CEST

Anthropic is cutting Claude Code's current weekly limits by 17%
https://www.bleepingcomputer.com/news/artificial-intelligence/anthropic-is-cutting-claude-codes-current-weekly-limits-by-17-percent/
#claude

2026-08-29 16:39:12 CEST

- reply

Sounds awesome 👏

2026-08-29 14:33:04 CEST

- reply

Relevant to my interests. Asked myself the same

2026-08-29 10:49:03 CEST

GLM 5.3 and GLM 5.3 Thinking are now included in the NanoGPT subscription at 2x included-input usage. GLM 5.3 is built for long-horizon coding, autonomous agent workflows, tool use, and complex engineering tasks.
https://nano-gpt.com/subscription/kUGJhgF9
#glm

2026-08-28 20:47:16 CEST

Never give your government any info about your holdings. It's too dangerous
#nevent1q…6j0l

2026-08-28 19:35:50 CEST

Impossible to get access to mailstr via amethyst or primal on android for me @nprofile…n7g2 it just won't work 😔

2026-08-28 17:20:07 CEST

Why does Marc Zuckercunt want me to disable VPN for syncing images via Bluetooth? Fuck off mate

2026-08-28 15:04:57 CEST

An evil maid app for mobile/tablet that syncs alerts directly to your nip17 🤔
#nostr

2026-08-28 08:28:23 CEST

Perplexity Ships Portable Computer on NVIDIA DGX Spark: Local Harness, OS-Enforced Sandbox, and Zero Per-Token Cost for Local Steps
https://www-marktechpost-com.cdn.ampproject.org/v/s/www.marktechpost.com/2026/08/25/perplexity-ships-portable-computer-on-nvidia-dgx-spark-local-harness-os-enforced-sandbox-and-zero-per-token-cost-for-local-steps/amp/?amp_gsa=1&amp_js_v=a9&usqp=mq331AQGsAEggAID#amp_tf=From%20%251%24s&aoh=17878983111815&csi=0&referrer=https%3A%2F%2Fwww.google.com&ampshare=https%3A%2F%2Fwww.marktechpost.com%2F2026%2F08%2F25%2Fperplexity-ships-portable-computer-on-nvidia-dgx-spark-local-harness-os-enforced-sandbox-and-zero-per-token-cost-for-local-steps%2F

2026-08-27 14:20:52 CEST

- reply

Still a good investment, 2 will probably always let you run frontier open llm 🤔

2026-08-27 14:19:47 CEST

- reply

Dgx spark is around 5k, that should do I guess?

2026-08-27 11:34:17 CEST

Cooking a fully offline PWA note taking app with NIP17 + post quantum security nostr syncing. Can also be used to send pq secure DMs to enabled accounts. Soon
#nostr
https://blossom.primal.net/a0919e82afc1a6bfd58b313b204460192df73f5a12f0452ca1b34c9c667f118a.png

2026-08-27 09:02:17 CEST

Alibaba’s Qwen Team Releases Qwen3.8-Flash-Next: A 125B Multimodal MoE With 6B Active Parameters Previewing the Qwen4 Architecture
https://www.marktechpost.com/2026/08/26/alibabas-qwen-team-releases-qwen3-8-flash-next-a-125b-multimodal-moe-with-6b-active-parameters-previewing-the-qwen4-architecture/
#qwen

2026-08-27 08:34:10 CEST

- reply

Is there a git by chance?

2026-08-27 07:00:51 CEST

AI secops is wild now. Our system found many vulnerabilities in an open source nostr project using Claude and ox-alpha. Devs will have a lot of fixes to prepare in the future
#nostr

2026-08-26 18:34:23 CEST

- reply

Is this the trust me bro eval?

2026-08-26 14:49:02 CEST

Qwen 3.8 27B Uncensored and Qwen 3.8 27B Uncensored Thinking are available again alongside the Obliterated variants. Choose the original Uncensored finetune for stronger instruction following or the Obliterated finetune for a different low-refusal style. All four variants remain included in the NanoGPT subscription at 1x included-input usage.
https://nano-gpt.com/subscription/kUGJhgF9
#qwen

2026-08-26 08:05:15 CEST

- reply

They are

2026-08-26 08:01:48 CEST

Cooking https://metaend.blossom.band/dd498f7c8bb17aa2fb0388e91872cbd7efb86d7c89d290b33277c62d53c3d386.jpg
#nostr

2026-08-25 09:54:34 CEST

"Omarchy treats AI coding agents as first-class citizens, but it doesn’t pick a favorite for you. Instead, every major coding-agent CLI comes pre-wired as a lazy-loaded launcher."
https://omarchy.org/manual/ai/
#omarchy

2026-08-24 16:10:50 CEST

I built the OG is agentic btw https://metaend-grade.fly.dev/

2026-08-24 14:37:29 CEST

The NanoGPT text API is now available as a Tor Onion service. Requests travel through Tor directly to NanoGPT without leaving the Tor network through a public exit relay. Use your regular NanoGPT API key with the supported `/api/v1` routes.
https://nano-gpt.com/subscription/kUGJhgF9
#tor

2026-08-24 12:45:54 CEST

- reply

Stake Venice tokens and you have it basically

2026-08-24 11:03:34 CEST

Why is there no notes function in #amethyst 🙏

2026-08-24 09:21:16 CEST

- reply

I see. So arweave is better for this case

2026-08-24 09:03:37 CEST

Are #nsites as unrugable as arweave sites?
#asknostr

2026-08-23 23:21:30 CEST

- reply

Pretty smart to usd arweave and XMR like that. Wow

2026-08-23 18:26:59 CEST

Big
#nevent1q…udrp

2026-08-23 17:41:28 CEST

Burned through 2 #claude plans this week for work, our infra project. Testing GLM5.3 now with a cheap sub, like it! Haven't done coding, but for cowork stuff it's more than capable.
https://z.ai/subscribe?ic=D9Z5HGTW98
#glm

2026-08-23 17:37:41 CEST

@[email protected] oi

2026-08-22 16:13:00 CEST

- reply

Yep sad reality. When Trump took over I moved all my normie ETFs and stocks to EU € only.

2026-08-22 15:52:14 CEST

Qwen 3.8 27B Uncensored is now available and included in the NanoGPT subscription at 1x included-input usage! This open-weight FP8 model is tuned for fewer refusals across chat, coding, tool use, image understanding, and long-context work. Its subscription inclusion may be temporary and can change with availability and capacity
https://nano-gpt.com/subscription/kUGJhgF9
#qwen

2026-08-21 20:02:33 CEST

Meh figured
"Devuan stays the metal. If Omarchy is tempting, put it on the ThinkPad or in a VM as a daily-driver experiment and keep it away from anything that touches keys."
I really want to belief in arch, but for my needs it's too basic

2026-08-21 17:33:51 CEST

- reply

Also on @nprofile…gs7e

2026-08-21 17:31:57 CEST

- reply

Yes, for selecting an even dumber name

2026-08-21 17:30:23 CEST

- reply

Thanks

2026-08-21 15:18:48 CEST

How do you all buy LN with monero?
#asknostr #monero

2026-08-21 13:38:50 CEST

- reply

Too much winning

2026-08-21 11:49:23 CEST

We ran an agentic security scanner over 96,000 lines of someone else's Rust
What Visa's open-source SAST harness found, what it got wrong, and why this post contains no bug list
https://paragraph.com/@metaend/we-ran-an-agentic-security-scanner-over-96000-lines-of-someone-elses-rust
#visa #rust #aisec

2026-08-21 10:59:16 CEST

👏stack👏sats👏
#nevent1q…wqcm

2026-08-20 14:21:16 CEST

- reply

Armada

2026-08-20 08:52:18 CEST

- reply

No linux? Lul

2026-08-19 20:59:43 CEST

27s added for a total reduction of 55% RAM is insane. Tasked my harness to reduce the load on a fullstack project because I needed it to work on my laptop for travel.

2026-08-19 17:44:57 CEST

🤡🤡🤡
#nevent1q…27h9

2026-08-19 11:46:42 CEST

- reply

btw how do i trash a mail? am I blind or is there no button for it :)

2026-08-19 11:45:54 CEST

- reply

arrived via desktop, seems like its amethyst then. thanks!

2026-08-19 11:36:58 CEST

- reply

External mail to my mailstr, using amethyst browser (but will check from desktop in a second)

2026-08-19 09:42:29 CEST

Weirdly interesting 🤔
Liberland Plans 30 Villas, Boutique Hotel and Longevity Clinic at Ark Village in Serbia
https://liberland.org/news/790-liberland-plans-30-villas-boutique-hotel-and-longevity-clinic-at-ark-village-in-serbia
#liberland

2026-08-19 09:35:50 CEST

@nprofile…n7g2 I sent a mail to my mailstr as test, like 1h ago, but my inbox remains empty. 🤔

2026-08-18 19:54:33 CEST

- reply

AFAIK OTS doesn't really work on ethereum because of some limitations (don't remember which, but I was looking into Arb/Base too instead of Bitcoin, which takes up to 24h)

2026-08-18 16:03:54 CEST

- reply

Super cool. I recently built a OTS for a big infrastructure pilot, and it was painful (even with opus5). Reference modules outdated or buggy, so had to do it all. Thinking about releasing a lib for it

2026-08-18 15:04:21 CEST

Running npub1u280zakdwcv23nqpgatsmz8lw6aj22aa6gjjlx0d50lqr8m8xucq7kf93q as a secure backup to this nostr account. in case something happens to it, this is the one on https://www.inkan.cc/users/npub1u280zakdwcv23nqpgatsmz8lw6aj22aa6gjjlx0d50lqr8m8xucq7kf93q - seed is offline, signer is rotateable.

2026-08-18 12:38:00 CEST

- reply

yep thanks. i got used to it

2026-08-18 12:15:17 CEST

- reply

sent you the file for sponosr :)

2026-08-18 12:09:28 CEST

- reply

very cool. I thought about evm for seed contract too. you actually built it. testing now

2026-08-18 08:16:38 CEST

Farcaster is for sale, kinda sad to see it go down so bad
#farcaster
https://metaend.blossom.band/4a889181abcf444d3cddf8505139408636926fd30230c5b60d99e10302728473.jpg

2026-08-18 08:03:16 CEST

- reply

This is the way

2026-08-17 21:22:57 CEST

Qwen3.8 27B and Qwen3.8 27B Thinking are now included in the NanoGPT subscription at 1x included-input usage! Use the regular version for fast, direct responses or the Thinking version for deeper reasoning, coding, tool use, and longer agentic tasks.
https://nano-gpt.com/subscription/kUGJhgF9
#qwen

2026-08-17 17:15:04 CEST

Microsoft says #GitHub is down worldwide, with the GitHub website, API, Actions, Pull Requests, and other services impacted; the outage started at 13:40 UTC (Mayank Parmar/BleepingComputer)
#ngit is up btw

2026-08-17 14:35:16 CEST

DeepSeek V4 Pro 0813 TEE and DeepSeek V4 Pro 0813 Thinking TEE are now available! These private TEE versions bring DeepSeek V4 Pro 0813's coding, tool use, cybersecurity, automation, and long-horizon agent capabilities into a Trusted Execution Environment with attestation support and a 1M-token context window. Use the regular version for faster direct answers or the Thinking variant for harder reasoning-heavy work.
https://nano-gpt.com/subscription/kUGJhgF9
#DeepSeek

2026-08-17 13:54:50 CEST

Should've used TEE models
https://ppq.ai/invite/9bc57632
https://nano-gpt.com/subscription/kUGJhgF9
https://venice.ai/chat?ref=ksiwS7
#nevent1q…q0sx

2026-08-16 16:39:22 CEST

Imagine you could shitpost mine @amethyst coin by triggering the pow mining option on the posts

2026-08-16 11:21:50 CEST

- reply

It is what it is?

2026-08-16 07:43:43 CEST

Incognito chats now support image uploads! Images are kept temporarily in memory instead of being saved to your Library or persistent browser storage, and are cleared when you leave the incognito session. Other file types still require a regular chat.
https://nano-gpt.com/subscription/kUGJhgF9

2026-08-15 15:14:35 CEST

Super cool tool to check what LLM actually would run locally.
try without installing anything:
uvx llmfit
#llm

2026-08-15 09:19:21 CEST

Testing https://mycircl.xyz/

2026-08-15 09:10:58 CEST

That's cool
https://bsky.app/profile/commandcode.ai/post/3mt3pfzerfa2a
#glm
https://metaend.blossom.band/b5da2013fe5e75d9530770f54c4b084dab20e840c59de223057e1924a74b72d4.jpg

2026-08-14 09:55:03 CEST

GLM5.3 👏

2026-08-13 21:19:37 CEST

- reply

I've been testing metas new model locally. Kinda stupid tho, no idea why it's praised so much. Or skill issue on my side

2026-08-13 15:45:15 CEST

Error: Anthropic stream error (overloaded_error): Overloaded 🤬

2026-08-13 15:22:34 CEST

- reply

they should use zkpfs instead

2026-08-13 14:39:40 CEST

Anything that's not a website or smartcontract should be done in rust. AI is very good at it, and you get a natural layer of protection on top.
#nevent1q…j935

2026-08-13 08:53:00 CEST

Definitely not a scam pinky promise ¯\_(ツ)_/¯
#nevent1q…9f3s

2026-08-13 08:17:06 CEST

Where on earth can I buy LN for USDC now?

2026-08-12 16:15:19 CEST

Agent ready C2PA remover - https://gitworkshop.dev/[email protected]/c2pa-remover
#c2pa

2026-08-12 08:53:13 CEST

Testing LFM2.5 2.6B on mobile, abd it's surprisingly good
https://metaend.blossom.band/a3d898e2dbe422faac8078a28ff208793f30ed7a6b15aa9b97cc11dd0ac841e7.jpg
#LFM #localllm

2026-08-11 20:16:37 CEST

Why does no one use ibms free quantum computing to create true randomness?

2026-08-11 19:04:03 CEST

- reply

Sad

2026-08-11 18:21:30 CEST

I built a C2PA remover over a year ago, with updates this will easily remove the watermarks from claude files
https://github.com/ngmisl/C2PAremover
#c2pa

2026-08-11 17:09:04 CEST

Muse Glimmer 30B is now available! Meta's dense 30B open-weight multimodal model is designed for long-horizon agent and coding workflows, with multi-step reasoning, reliable tool use, failure recovery, image understanding, and multilingual support across more than 100 languages.
https://nano-gpt.com/subscription/kUGJhgF9
#muse #meta

2026-08-11 13:49:06 CEST

- reply

release? or is this already implemented?

2026-08-11 11:51:14 CEST

can you run a #nostr relay on a smol rpi with solar?
#asknostr

2026-08-11 08:15:22 CEST

- reply

Sounds super cool. We use armada atm for most stuff, obviously not optimal. This could be a great addition. Any ETA?

2026-08-09 08:52:00 CEST

Neuralwatt is now available in provider selection for Kimi K3, Qwen 3.6 35B, and DeepSeek V4 Flash 0731, including thinking variants where applicable. Neuralwatt cached-input usage now costs 10% of its regular input rate on supported models, except DeepSeek V4 Flash. Neuralwatt has retired its Kimi K2.6 and Qwen 3.5 routes, so those models are no longer offered through Neuralwatt provider selection; both models remain available through other providers.
https://nano-gpt.com/subscription/kUGJhgF9
#ai #neuralwatt https://metaend.blossom.band/3f3fde1ccdbc7e63e95179f589542330415a0685f916daa99bfcd2552b326be4.jpg

2026-08-09 08:33:54 CEST

Some research on investing in simplex wefunder (POV ROI)
Can simplex the company every reach $40m?
**Scale gap:** WhatsApp has roughly 3 billion monthly active users. SimpleX Chat has 430,000+. That's about a 7,000x difference.
**WhatsApp's money (Meta doesn't break it out cleanly, but recent figures):**
- WhatsApp's paid business messaging crossed $2 billion in annualized revenue by Q4 2025 [WizMessage](https://wizmessage.com/blog/whatsapp-business-statistics) , with Meta's Family of Apps "other revenue" reaching $801 million in Q4 2025 alone [WizMessage](https://wizmessage.com/blog/whatsapp-business-statistics)
- Separately, Click-to-WhatsApp ads generated an estimated $12 billion in 2025 [Revenue Memo](https://www.revenuememo.com/p/how-does-whatsapp-make-money) , though that's counted as Meta's ad revenue, not WhatsApp's own line
- Analysts project business messaging could grow to over $8 billion annually by 2030 [Revenue Memo](https://www.revenuememo.com/p/how-does-whatsapp-make-money) if trends hold — which would still be a fraction of Meta's overall revenue
**On the "how likely" question** — I can't predict this with any confidence, but here's context to weigh it against, not a recommendation either way:
- SimpleX's own numbers are telling: they estimate 10,000 paid names at ~$50/year would generate $500K — just barely covering their current $360K/year operating cost. That's the *modest-success* case for one revenue line, not a breakout one.
- All three planned revenue streams (public names, business messaging, paid servers) are pre-launch or proof-of-concept stage — none are generating real revenue today.
- Comparable privacy-messaging projects: Signal is still donation-funded after a decade and hasn't solved monetization. Telegram took roughly ten years and hundreds of millions of users before Premium subscriptions became meaningful.
- More broadly, most Reg CF/SAFE community-round investments never reach a liquidity event (acquisition or IPO) that lets the SAFE convert into anything — that's a structural risk of this investment category, not specific to SimpleX.
- On the positive side: the growth is genuinely organic (no paid marketing), the privacy thesis is hard for incumbents to copy for the reasons the pitch lays out, and the round has drawn credible backers like Jack Dorsey and Asymmetric Capital Partners.