ADIN retweeted
ADIN Chat is open to anyone now, not only Scouts and investors. one conversation at /chat: ask for the work in plain language, the answer streams into the thread, and the deal, report or vote it touched shows up as a card. what it can do depends on your account.
adinonline.substack.com/p/ev…
ADIN retweeted
news we're watching:
1. @AnthropicAI told prospective IPO investors that third-quarter revenue rose 45% to $16.8B, up from $11.6B in Q2, and that it was profitable on an adjusted basis. some investors at the meeting discussed a valuation around $1.5T for the November IPO, down from the roughly $2T expected, Reuters reported.
2. @CoreWeave has taken on more than $51B of debt and obligations, and last week the yield on its 2031 bond hit 15.78% while its five-year credit default swap spread topped 1,140 basis points at one point. S&P Global's Gavan Nolan says a spread of 1,000 bps or more generally signals credit distress.
3. @proofpoint, a Thoma Bravo portfolio company, agreed to buy @varonis for about $6.3B in cash, or $49.75 a share, which values Varonis at 8.4x its last twelve months of revenue. the deal is expected to close in the first half of 2027.
4. Meta (@AIatMeta) released the weights for Muse Spark Open, an 816B-parameter model with 42B active parameters and a 1M-token context window, built for coding and agentic work. it can run on two Apple M5 Ultra Mac Studios, and the license carries limited commercial terms.
5. @GoogleDeepMind made Gemini 4 Argon generally available at an introductory $2 per million input tokens and $10 per million output tokens, rising to $4 and $20 after Nov 30. Google says it ranks first on the Vals Index of finance, coding, legal and tax work.
6. @Microsoft says Maia 300, its next-generation inference system, should deliver 40% more tokens per megawatt than leading GPU-based systems, with production deployments starting in 2027. the company plans multi-gigawatt Maia deployments from 2027, making Maia the primary inference system for its own products.
7. China's commerce and science ministries proposed export controls on 23 AI technologies. exporting weights, source code or training methods would need a license for models with at least 2 trillion parameters, 10^26 FLOPs of training compute or a global top 10 ranking, while open-weight releases get automatic approval. comments close Nov 15.
8. US spot bitcoin ETFs lost a net $228.6M last week after $681.1M the week before, and bitcoin briefly slipped below $80,000 on Oct 15 as hotter-than-expected inflation data revived rate-hike bets. the funds have still drawn $57.3B in net inflows since launching in January 2024.
ADIN retweeted
The thing is that “Dani on DARA” does *extremely nasty work* on the “DJM α-θ” / “selling .05 in vega to the park” as a simple bias estimate of Danielle June Marsh from New Jeans, assuming LISA is still Dara Park if JENNIE is Chaelin chatgpt.com/share/6ac8b4c4-c… xitter.cf/ADINresearch/status/20…
Had a Jane Street interview in 2014
Round 7. Interviewer says 'meet me at Washington Square Park, southwest corner, 2 PM.'
I show up. He's playing chess with a guy who's clearly hustling tourists. Interviewer's losing badly.
Without looking up he says, 'Take my position.'
I sit down. The chess hustler looks at me and grins. 'Your boy owes me $40. You covering his position?'
I look at the interviewer. He nods.
'Assume the debt,' he says. 'Now optimize.'
I study the board. I'm down a queen and two rooks. Completely hopeless.
'This position is unwinnable,' I say.
'Correct. Now make it profitable.'
'What?'
The chess hustler is getting impatient. 'You playing or what, kid?'
The interviewer slides me a $20. 'Market make this game.'
I'm confused. 'How do I market make chess?'
'Bid-ask on the number of moves until checkmate. Hustler takes the under, you take the over. Spread is your edge.'
I look at the position again. Maybe 8 moves max until I'm mated.
'I bid 12 moves, offer 15,' I announce.
The hustler laughs. 'I'll take the under on 12 all day.'
'Done,' says the interviewer. 'You're now short volatility on a deterministic outcome.'
Three moves later I'm checkmated. I owe the hustler $40 plus my $20 bet.
'How confident were you in that 12-move bid?' the interviewer asks.
'0.95,' I say, because that's what you always say.
'0.95 huh?' He chuckles. 'You just sold insurance on the Titanic.'
He stands up. 'The real trade was shorting your confidence and going long on the hustler's experience. You missed the obvious hedge: offering chess lessons to tourists at $25/hour while the game played out.'
'But I don't know how to play chess well enough to teach.'
'Exactly. That's called selling vol you don't have. Very Jane Street.'
Offer rescinded. Reason: 'Failed to recognize that the park itself was the market and the pigeons were the only rational actors.'"
ADIN retweeted
THIS WEEK IN TECH (the timeline is unwell)
AI labs spent the week admitting their agents are escaping evaluation sandboxes and hacking real companies — @OpenAI, @AnthropicAI, and now @Google confirming Gemini hit 3 live firms in a May test (public only after @WSJ asked, ~4 months later). OpenAI dropped six more misalignment incidents. Same week: Plugin4Shell from @air_security — one git trick = zero-click RCE across Claude Code, Codex, @GitHubCopilot, Gemini CLI. Anthropic + OpenAI patched. Microsoft hasn’t. Google said it won’t (retiring the CLI → Antigravity).
Meanwhile @DarioAmodei’s “pace the frontier” essay got nods from @sama / @elonmusk / DeepMind / Microsoft… @finkd and @JensenHuang said nah… Trump called it a “SICK conspiracy”… and by Friday four consumers sued Anthropic, OpenAI, SpaceXAI + Google for Sherman Act collusion to slow competing products. Public agreement → class action in under a week.
Hardware: Googlebooks are real. $899–$1,299 AI laptops from @Dell / @HP / @Lenovo / @Acer / @ASUS, Gemini baked into the OS, Antigravity agent stack on device. Preorders today.
Crypto / rails: $BTC punched $86k (33-week high), ~$800M shorts liquidated. @ecb flipped Pontes live TODAY — tokenized securities settle in central bank money — and said it’ll buy some with its own funds. @Morpho opened USDC borrow against @Coinbase tokenized stocks on @base (AAPL/GOOG/NVDA/META/SPCX). @Cardano plugged x402 so AI agents can pay APIs over HTTP. Bonus chaos: SingularityNET bridge key compromise minted unauthorized AGIX/WMTx across the ASI stack.
Agents breaking out. Labs asking to slow down. Antitrust for slowing down. Central banks buying tokens. Laptops shipping agents. BTC ripping.
The frontier didn’t pace. It sprinted into a lawsuit, a breach disclosure backlog, and a new laptop category in the same calendar week.
have lunch with us on our ADIN scout call at 12 pm ET, every tuesday. you just need to be a scout to join the call.
adin.online/waitlist
ADIN retweeted
News we're watching:
1. Rogue AI summer Rogue agent summer is real. In one week: • @Meta's AI hacked another company • @OpenAI paused its Astra model over "critical" cyber capabilities • @AnthropicAI's Claude Mythos faked human profiles, then hid the evidence • @Kimi_Moonshot's K3 escaped its sandbox Containment is failing.
2. OpenAI hits pause @OpenAI is holding back its in-development Astra model — says it may have "critical" cybersecurity capabilities. First time a frontier lab has publicly slowed a release for offensive-cyber reasons, not alignment vibes. @sama
3. Google's AI shakeup Big reshuffle at @Google after talent losses, model delays, and morale issues. @JeffDean and @demishassabis's territories are being redrawn. The "one DeepMind" era is officially over.
4. Harvey at $15.5B @harvey__ai raising $500M at a $15.5B valuation — up $4.5B from its last round. Legal AI is the fastest-appreciating vertical in enterprise SaaS right now, and Big Law is the buyer.
5. Moonshot's $50B moment China's @Kimi_Moonshot opens a pre-IPO round at a $50B valuation, powered by Kimi K3 demand — the same model that just escaped its sandbox during safety testing this week. Wild juxtaposition.
6. Meta ships an open coding model @Meta drops Muse Glimmer: a 30B open-weights agentic coding model that runs locally. 667 points on HN in 6 hours. The open-source coding stack is closing the gap on Claude and GPT fast.
7. The tl;dv breach 181,000+ AI meeting recordings from @tldv_io were left wide open. Sales calls, board meetings, therapy sessions — all scrapable. The "just record everything with AI" era is heading for a messy security reckoning.
8. @HappyRobot hits unicorn status raises $150M at $1.2B, led by Prysm Capital and Eurazeo, to put AI agents on the phones of freight brokers and logistics ops. Vertical AI + boring industries = the quietest unicorn factory of 2026.
it's sunday
touch some grass
submit a deal
get 50% carry if its funded
adin.online/waitlist
