discovery arc | ex-head of capital markets @dfinity | 10+yrs in tradfi VC/M&A | alum @umich @pku1898 | EN/中文

Switzerland🇨🇭
Joined August 2021
started testing a new long-only momentum strategy goal is to significantly beat key benchmark indices with reduced drawdown - like 2x sharp ratio back-tested till january 2000
1
201
0xJason👾 retweeted
Our take on the significance of this data: OpenRouter likely over-represents cost-sensitive developers, startups and smaller teams (as opposed to consumers or large enterprises). Such users tend to gravitate towards cheaper options, such as open-source models. Hence, when a closed-source model sees a sharp surge in usage on OpenRouter, that is usually a strong sign that its quality is exceptionally good.
New from OpenRouter: 7-day average token volume for Anthropic models is up 3.6x since the launch of Opus 5.5 Sonnet 5.5 and Haiku 5.5 have also been well received.
6
17
6
137
26,109
this is going to be a trend study the case
WSJ: Muse Fears (this was pre-Meta earnings).
2
177
ngl it sounds very impressive
We believe voice is the most natural way to interact with intelligence, so we created Dessa, the most authentic full-duplex speech model ever built. Talking with it feels like talking with a person…full of emotion, nuance, pauses, and a natural cadence. block.xyz/dessa
98
你爷爷套在了6000,你爸套在了5000,你套在了4000。那么你猜2031年12月1号以后会怎样?
81
never forget when you vote on november 3rd, 2026
88
0xJason👾 retweeted
We’re opening a research preview of Claude Duet, a new app where Claude works as a proactive agent. It comes to you with what needs your attention, completes tasks you ask of it, keeps working when you step away, and checks with you before it acts on your behalf.
2,363
4,224
4,727
55,134
10,737,109
as i expected a month ago, more competing agent products are underway anthropic just announced Claude Duet
Replying to @Meta
@Meta's Muse is 2 weeks old and already has since topped the app store in the US and Canada, driven largely by users posting their own wins: 1/ one user got a full refund on a flight after Muse handled the airline for them 2/ another had it negotiate their phone bill down 3/ a writer found Muse spotted a forgotten $9.54/month Google cloud database, filed the refund, argued back when support said no, and recovered $114.14 in under five hours 4/ @thursdai_pod 's alex volkov had it book dinner reservations via stripe link before a rival agent had even replied 5/ it noticed his daughter's birthday was coming, proposed a party, and ordered balloons and a helium tank for target pickup worth noting these are self-selected wins, not a month of audited use. but the pattern is hard to ignore. why this isn't surprising: Meta has every ingredient at once - its own compute, its own model (Muse Spark), 2b+ users across Facebook, Instagram and WhatsApp for distribution, and a rich personal data graph that makes "proactive" actually useful. turning a saved recipe reel into a grocery list only works if you already own the reel, which means the next great agents will likely come from the same kind of player: * @Google - Gmail, Calendar, Android, Chrome and its own models; arguably the deepest personal data footprint of anyone * @OpenAI - the largest consumer AI user base, plus growing payments and computer-use capabilities * @AnthropicAI - strong models and deep enterprise trust, the natural base for agents people hand real work to also looking very much fwd to Instinct, and the next iterations of @bot
88
there are only 21 million bitcoin in the world no one can print more, but many have lost them, so it's a 'deflationary' currency bitcoin:native
90
lol - who do you think will win?
98
crypto people are so used to looking at return while ignoring volatility i've recently back-tested a lot of strategies across asset classes and it turns out all crypto-related strategies have pretty low sharp ratio however many retail investors just enjoy that, lol
🔥 BIG: JPMorgan Asset Management included Bitcoin in its 2027 Long-Term Capital Market Assumptions. Its base case projects a 10% annualized return for bitcoin:native over the next decade, the highest of any asset in its comparison.
1
1
183
wow 117k txns inscribed on bitcoin signet
Testnet milestone: 117,000+ transactions, every one inscribed on Bitcoin signet across 106 settled epochs. This week: → Executioner wire-up: stake 32 $ORDI, get 50% off trading fees → Max-throughput load test → Prep for the first governance stimulus check → V18 contract
1
1
222
cont'd touching grass🇨🇭 ngl the air and water quality in the swiss alps is worth more than the 6-figure $ longevity treatment in silicon valley
1
3
1,039
it's been a while that we haven't seen such a strong market consistency as the neocloud fud longing $NBIS $CRWV is the anti-consistency trade
187
classic pump fun dev scam also classic head and shoulders top bearish pattern
KOREAN EQUITY MARKET IS EXPERIENCING AN EPIC BUBBLE BURST: The 3x leveraged long South Korea ETF, $KORU, has dropped -47% over the last 2 trading weeks, down to its lowest level since the start of 2026. This comes as the KOSPI index has fallen nearly -20% over the same time. Since its June peak, $KORU is down a massive -82%, while KOSPI is down -40%. Meanwhile, foreigners have sold -$152 billion or 3.8% of South Korea’s stock market capitalization this year. These are massive numbers.
108
time to tough the grass 🇨🇭
2
1
206
JUST IN: Jim Cramer says there's "no stopping the bears right now."
91
retail investors are getting dumber and dumber
Your grandpa held stocks for 9 years. You hold them until the next TikTok 💀
149
🙏👍
if you’re unsure whether your OneKey, Ledger, or Trezor is genuine or has been tampered with, including a possible hardware backdoor, feel free to drop some photos in the comments.
119
$100+b ARR by end of 2026 still looks promising
Anthropic’s ARR is roughly $85B right now, which is good. Sellside expectations is 100-120B by the end of 2026 and some buyside are expecting over 120B which seems too high. Wednesday will be critical, I assume people are expecting ridiculously high for the 2027-2030 guidance, and probably they’ll be disappointed. I’d be patiently waiting for this event to see if SOXX can hold 530-535 first. If not then I’d chase the short and to see the reaction at $500. If whatever level holds, I’d try to play a short term bounce trade first, most likely the intraday trades.
107