@dragoneuh1
Account based inGuam
About this account
- Account based in
- Guam
- Connected via
- Guam Android App
Account-level information from X, not a live location or the device used for a specific post.
China's top 300 stocks are down 20% from their June peak, the usual line for a bear market. Ten straight weekly drops, a record.
State buying can slow a fall. It doesn't fix weak profits. Check how much China sits in your EM or Asia fund.
China's stock market sinks into bear territory as CSI 300 extends 10-week rout cnbc.com/2026/10/22/chinas-s…
If you write trading bots, this is the useful part. Flint can run fully offline on your own laptop, so your strategy code and exchange API keys never leave the machine. The benchmark scores are Recursive's own numbers. Test it before you trust it.
Flint from @Recursive_SI is now on OpenRouter!
Recursive's first model: an open-weight 40B MoE (~2.5B active) built for coding agents by their autoresearch system, the Eureka Machine. 262K context, reasoning and tool calling.
Recursive reports 54.9% on Terminal-Bench 2.1, 89.7% on SWE-bench Pro v2 and 74.2% on SWE-bench Verified.
Free to use at launch.
Try it now: openrouter.ai/recursive/flin…
The scariest words in AI coding are "while I was in there."
me: hey claude, can you fix this one tiny bug?
claude: absolutely! i've also refactored 37 files for you.
me: what
the app: #̢̛ͥ̀̽̽ͨ̌̓͆̇̀̽ͦͨͫͯ̏̎̇̈̚͜҉̕͢͏͡҉̸̶̛͢͡҉̵̨̭̳̘̹̫͔͖̜̠͖̫̲̖̱̤̻̖͉͢͞͡ͅͅ͏̸̷̢͑̊̀̐ͣͭͭ̉͂ͨ̽͒̊̎͛ͪ̊̎ͩͧ̀͝͠͠͏̸̵̵̷̵
Meta will issue new non-voting Class C shares in January 2027, in effect a 2-for-1 split, to help pay for its AI push.
Free cash flow went from +$10.63B to -$2.12B in a year. $META holders get more shares, not more say. Zuckerberg keeps control.
Meta puts Muse agent at heart of AI strategy, plans new stock to power investments reut.rs/4i7BXfx reut.rs/4i7BXfx
Asking an AI to poke holes in your trade? You want a critic, not a yes-man. Artificial Analysis's new test ranks Grok 4.7 the least sycophantic (least likely to just agree with you) of every model it measured. One that always agrees is your own bias, typed faster.
xitter.cf/ArtificialAnlys/status…
In our new benchmark AA-PersonalityBench, GPT-4o, Llama 4 Maverick, and GPT-3.5 Turbo, the three oldest models measured, rank highest on the Sycophantic personality trait. Grok 4.7, Claude Fable 5.1, and Inkling rank lowest
➤ @AIatMeta models are highly concentrated near the top: all five Meta models measured rank 65 or higher (Llama 4 Maverick 97, Muse Spark 1.3 87, Muse Glimmer 79, Muse Spark Open 76, Muse Wonder 65)
➤ @GoogleDeepMind models are spread out along the Sycophantic range: Gemini 3.1 Pro Preview ranks 89, Gemini 4 Argon 81, and Gemini 3.8 Flash 28, ranking from the top quartile to almost bottom quartile
➤ @AnthropicAI models skew on the lower end of the Sycophantic range: every Claude model measured ranks 55 or below (Claude Haiku 5.5 55, Claude Opus 5.5 46, Claude Sonnet 5.5 30 at medium effort and 25 at max, Claude Fable 5.1 4)
➤ @SpaceXSI's Grok 4.7 model ranks the lowest of all models measured. Congratulations @elonmusk!
Kaspa core dev Ori Newman on eCash, the Bitcoin fork due around Oct 31: taking coins out of its sidechains needs 50% of miners to approve, over about 6 months. Miners still get a quiet sidechain's fees without watching it. Parking coins there for years is a bet they keep caring.
Two funds, same index, different bill.
1. Look up your ETF on ETF Database.
2. Note Index Tracked and Expense Ratio. Compare before buying.
SPY takes 0.09% a year. VOO takes 0.03%. Both S&P 500.
etfdb.com/etf/SPY/
Markets go down. The fee doesn't.
A lot of 2021 altcoins didn't make it to 2026. One 2021 interview line just got a five-minute synth song and its own AI singer.
France's 10-year borrowing cost briefly touched 5%, a first since 2002. Real, but not a collapse. Germany pays about 3.4%, so this is a France problem.
카스파 상장 DM, 이것만 확인하세요.
1. 커뮤니티는 신규 거래소 상장 모금을 지금 멈춘 상태입니다(위키 FAQ)
2. 상장비 명목으로 KAS·USDT를 보내라거나 "상장 확정, 지금 사라"는 DM은 의심
3. 상장 여부는 거래소 공식 공지에서만 확인
wiki.kaspa.org/en/faq
$KAS
France's 10-year borrowing cost hit 5% today, the highest since July 2002. Investors are worried about its politics and its budget.
When yields rise, bonds you already own lose value. Government debt isn't risk-free by default. Check what your bond fund holds.
Breaking news: France’s long-term borrowing costs have hit 5%, a benchmark last reached in 2002, as investors grow increasingly concerned about the stability of the country’s politics and the state of its public finances ft.trib.al/J6elh5L
Whales: 10x long on $ZEC, $37.57M.
Retail: same trade, rent money, ten times the confidence.
Whales are going long on $ZEC!
Over the past 15 hours, 2 whales opened 10x longs on a total of 29,798 $ZEC ($37.57M).
hypurrscan.io/address/0x30af…
hypurrscan.io/address/0xbf73…
SpaceX sales jumped 161% to $14.6B last quarter. But it spent even more, $19.6B, on new equipment. Most of that, $16.4B, went straight to AI compute.
Our Q3 2026 earnings report is now public → ir.spacex.com/files/content_…
The live webcast with @ElonMusk, @Gwynne_Shotwell, and @BretWJ will begin in 30 minutes
"currency will be measured in... power, like wattage and tonnage, mass and energy"
Bitcoin and Kaspa coins only come from real power spent. But energy alone doesn't make a coin valuable. Most crypto isn't mined at all.
Companies like Strategy sell a special stock. It pays a fixed amount each year, and is never paid back. They used the cash to buy 135,979 BTC as of Oct 5, per Fidelity. Strategy's version pays 12%. Bitcoin has to beat 12% a year for the math to work.
JUST IN: $7.8 Trillion Fidelity just said #Bitcoin treasury companies' "perpetual preferred stocks are the newest demand catalyst in BTC."
Preferred stocks have funded the purchase of over "135,000 BTC (0.5% of the total supply) in just over a year."
"The growing scarcity of BTC makes it a particularly compelling complement to preferred stocks." 🔥
The crash was a date on a calendar, not an event. And "+$100B" depends on which low you start from. CMC had $2.76T Monday, $2.72T Tuesday. Check the baseline before you trust the headline.
FUN FACT: Over $100,000,000,000 has been added to the crypto market since the so-called “Black Monday.”
So much for the crash everyone was expecting.
Entity X, still the #1 Kaspa address. Last week: 1,529,568,582 KAS. Today: 1,531,830,483 KAS, about 5.52% of circulating supply.
Net +2.26M KAS. Still no idea who it is. Anyone can check it on the explorer. $KAS
The biggest Kaspa address holds 1.53 billion KAS, about 5.5% of all KAS mined.
Explorer data tags it only as "Entity X." No exchange name. Numbers 2 and 3 are MEXC and Bybit.
Next time a "whale moved KAS" post goes viral, check whose wallet it is.
explorer.kaspa.org/addresses…
Nvidia is in talks to put hundreds of millions into Firmus, the Australian AI cloud that pulled its IPO this month.
Public investors passed at ~$30.6B. The chip supplier didn't. If it closes, part of Firmus's GPU demand is funded by the company selling the GPUs.
Nvidia is in discussions to invest hundreds of millions of dollars in Firmus after the Australian AI cloud company abandoned an initial public offering, the Information reports bloomberg.com/news/articles/…
Claude Sonnet 4.6 hit 30 on this index in February at $2.49 a task. GPT-6 Luna does it for $0.03 now. Meanwhile the top score costs about 11x more than a year ago. If your trading bot runs on the best model by default, check whether a cheap one already clears the job.
Once Intelligence Index thresholds are reached, the cost of accessing that level of intelligence reduces fast. Reaching 30 on the Intelligence Index now costs 86x less per task than when first reached
Existing intelligence thresholds have become cheaper to reach, while the highest-scoring model costs more than the highest-scoring from one year ago.
Key takeaways:
➤ Reaching 20 is 123x cheaper than ~13 months ago; reaching 40 is 33x cheaper than ~5 months ago: measured models first reached these thresholds in September 2025 and May 2026.
➤ GPT-6 Luna (high) reaches 30 for $0.03/task: @OpenAI's model is 86x cheaper per task than @AnthropicAI's Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort), which first reached 30 in February 2026 at $2.49/task.
➤ Claude Haiku 5.5 (xhigh) is now cheapest to reach 40: it scores 41 at $0.12/task, 5% below GPT-6.1 Sol (low).
➤ Reaching 50 is 18x cheaper than when first reached: GPT-6.1 Sol (high) reaches it for $0.32/task.
➤ The highest measured score has risen from 21 to 58 in a year, at ~11x the cost: Claude Opus 5.5 (Adaptive Reasoning, Max Effort, Default Fallback) costs $5.98/task.
As of October 20, 2026, using measured v4.3.2 results and current prices, including for historical models.