
Running bitcoin

Norwegian coder & systems architect. Developer of Mats3 and MatsSocket libraries: mats3.io Left-leaning Bitcoin maximalist.

Running bitcoin

Okay, this is super-strange. Jumping back in to that conversation, it AGAIN shows 96%. Very sticky, that wrong number..

Have it occurred to anyone else than me that there's another way to get the CPI down outside of CRASHING the economy?
FED MEETING MINUTES (11/18/26): 1. Almost all Fed officials saw “further action” appropriate 2. A couple Fed officials believe a hike of more than 25 basis points may be warranted in December 3. Some officials favored a 25 basis point rate hike in October 4. Returning inflation to 2% in a “timely manner” remains top priority 5. A few officials who supported no change in rates indicated the decision was a “close call” 6. A few officials believe inflation could prove more persistent than expected The path is set for a December rate hike.

After starting that chat window, it corrected itself eventually. But I've now seen this more than once - it starts out with some "you basically have nothing left", even though: Yes I do.

.@AnthropicAI you have a bug in the claude-ai web interface: It reports 96% used, while the usage-page of settings, in the same tab, shows 38% fable and 76% overall. I wonder if it doesn't take a "Fable reset" into account (of which I've used 1).

[Bitcoin] Dette er komplett vås: @barebitcoin er steinbra. Det denne tullingen snakker om, er en "fork" av Bitcoin som er verdt svært nær 0 - det er ikke Bitcoin.
Very different than @barebitcoin som stjeler sine kunders bitcoin uten å blunke. For the record: This Norwegian exchange @barebitcoin steals it's customers bitcoin. Stay away from them.

Wonder how much of all the dipping is due to the Anthropic IPO? Some? A little bit? None at all?

[Bitcoin] If you found Bitcoin a bit on the expensive side, and wanted to buy when it dipped .. .. well, this is it! 😁⏫

Interesting question indeed.
With all these cyberattacks happening across the USA, how come none of them have targeted the DOJ servers where the unredacted Epstein files are stored?

Hear, hear. I mean, all religions do it to degrees. Make children believe in some fictional Santa Claus that watch and judge you, distancing you from reality and strip away the ability reflect over life, independently. But JW is downright nasty.
Something that doesn’t get talked about enough is the negative social control and psychological abuse experienced by children growing up as Jehovah’s Witnesses. Society is afraid to challenge religion, and as a consequence we fail these children. It’s such a shame.

Okay, some chilling reflections from a guy that had preview. Astra 6.2 have some issues. Notice the mentions of Opus.
Theo - t3.gg @theo Nov 17 Why I hate GPT-6.2 Astra During early testing, I found myself so frustrated with this model that I spent a day writing a letter to OpenAI about it. They fixed a small handful of my complaints, but not enough for me to

[AI] Hah, they actually did it. Hours from the Anthropic IPO closing, OpenAI pull out their new, performance-cracking, cost-demolishing new model. Look at that poor chart, where the cost X-axis needs to fit Claude models, while GPTs are a vertical line near zero.. 😆
GPT-6.2 Astra—a frontier model built for your hardest projects. It sets new highs on evals across professional work and scientific research at less than a third of the cost per task of GPT‑6 Astra. You’ll see the difference in things like slideshows, websites, and 3D designs.

What an insane amount of drivel. And, is this anything Czech people are proud of? "In 1948, weapons and aircraft from the Czech Republic helped save the State of Israel. Today, Israeli defense systems help protect Czech skies. 78 years ago, you stood by our side. Today, we stand by yours."
אירחתי היום בירושלים את ראש ממשלת צ׳כיה, אנדריי באביש, ידיד אמת של מדינת ישראל. אמרתי לו דבר ברור: הטילים של איראן, הטרור שלה ושאיפותיה הגרעיניות אינם נעצרים בגבולות ישראל. אירופה אינה צופה מן הצד. אירופה היא מטרה. ישראל ניצבת בחזית המאבק של העולם החופשי נגד ציר הטרור האיראני. ריסקנו את מכונת הטרור של חמאס, שיתקנו את חיזבאללה, הלמנו בחות'ים ופגענו קשות בתוכניות הגרעין והטילים של איראן. המלאכה עדיין לא הושלמה, אבל כפי שהבטחתי – שינינו את פני המזרח התיכון. לצערי, יש מנהיגים באירופה שמעדיפים לגנות את ישראל במקום להילחם בטרור. אבל צ׳כיה בוחרת בדרך אחרת. היא עומדת לצד ישראל באומץ, בבהירות מוסרית ומתוך הבנה שהמאבק שלנו הוא גם המאבק שלה. אמרתי לראש הממשלה באביש: ישנם מנהיגים חלשים שמפנים את גבם לישראל, אבל אתה עומד זקוף. אינך הולך אחרי העדר, אינך נרתע בפני הטרור. אתה הולך בעקבות המצפון שלך. בשנת 1948 נשק ומטוסים מצ׳כיה סייעו להציל את מדינת ישראל. היום מערכות הגנה ישראליות מסייעות להגן על שמי צ׳כיה. לפני 78 שנים עמדתם לצידנו. היום אנחנו עומדים לצידכם. זוהי חברות אמת. ברוך הבא לירושלים. 🇮🇱🇨🇿

Perfect.
🇪🇺🇷🇺 Khallas suggests EU spy agencies steal Russian military records to push through visa ban on SMO veterans EU foreign policy embarrassment Kaka Khallas has floated the idea of European intelligence services pulling the personal data of Russian SMO participants from Russian state databases. She spoke at a press conference after the EU foreign ministers' meeting in Brussels on Monday. The goal is to make the planned EU entry ban on Russian veterans workable. Brussels wants the ban in its 22nd sanctions package, but several member states oppose it. They question how it could be enforced legally and in practice. Kallas had an answer for the doubters: "Russia is a very bureaucratic country. They keep records of all the people who took part in this war. So I think getting this data is fairly easy, given that everything is digital now… We all have very good intelligence specialists." In other words, the EU's top diplomat is publicly suggesting that member states break into another country's government systems to build a blacklist of its citizens.

[AI/Adjacent] This was interesting. Computer Vision in 2023, but intriguing wrt. whether AI might be an exponential or sigmoid. If you're in the early part of the curve, it is easy to over-extrapolate.
If you haven't seen, i highly recommend @jon_barron cvpr23 talk about this Slide: jonbarron.info/data/cvpr2023… Haven't searched for recording

Okay, AA was already there. This ain't no Opus 5.5 killer.
Alibaba has launched Qwen4 Max, gaining 9 points over Qwen3.8 Max on the Artificial Analysis Intelligence Index and landing 3 points behind Kimi K3.1 Pro at ~11% lower Cost per Task Qwen4 Max, using xhigh reasoning effort, scores 43 on the Intelligence Index, 9 points ahead of Qwen3.8 Max (xhigh) and 3 points behind Kimi K3.1 Pro (max, 46) Key takeaways: ➤ Qwen4 Max improves on every evaluation in the Index vs Qwen3.8 Max (xhigh, 34): HLE-Diamond +21 points (24% to 45%), AA-Omniscience +18 points, AutomationBench-AA +17 points, CritPt v2 +13 points, Terminal-Bench-Science +11 points, Terminal-Bench 4.0 +9 points, GDP.pdf +3 points, FrontierCode 2.0 +3 points and AA-Briefcase +82 Elo. ➤ Qwen4 Max costs $3.27 per Intelligence Index task, ~3x less than Qwen3.8 Max ($10.02) and ~11% less than Kimi K3.1 Pro ($3.68): This is driven by 25% lower per-token pricing than Qwen3.8 Max, 70% cheaper cache hits and lower token usage. Tencent Hy4 costs ~3x less per task ($1.03) than Qwen4 Max but scores 8 points lower on the Index. ➤ Qwen4 Max uses 39% fewer output tokens per task than Qwen3.8 Max: ~105k output tokens per Intelligence Index task vs ~171k for Qwen3.8 Max, ~72k for Kimi K3.1 Pro and ~78k for Tencent Hy4. Qwen4 Max uses fewer tokens, while increasing 9 points on the Intelligence Index ➤ Improvements in AA-Briefcase (knowledge work): On AA-Briefcase, our evaluation of agentic work across real business files, Qwen4 Max reaches 1686 Elo vs 1674 for Kimi K3.1 Pro and 1446 for Tencent Hy4. ➤ Qwen4 Max trails Kimi K3.1 Pro on Terminal-Bench 4.0 and Tencent Hy4 on FrontierCode 2.0: Qwen4 Max scores 48% on Terminal-Bench 4.0 vs 67% for Kimi K3.1 Pro and 37% for Tencent Hy4. On FrontierCode 2.0, @cognition’s private evaluation of real-world coding tasks, it scores 26%, level with Kimi K3.1 Pro (26%) and behind Tencent Hy4 (31%). Key model details: ➤ Context window: 1M tokens ➤ Multimodal: Text and image input with text output ➤ Pricing: $1.50/$4.50 per 1M input/output tokens on Alibaba's first-party APIs, with a $0.075 cache hit price. This is lower than Qwen3.8 Max across the board ($2.00/$6.00, with a $0.25 cache hit price) ➤ Availability: @Alibaba_Qwen 's first-party APIs