E268
Anthropic's $30B Ramp, Mythos Doomsday, OpenClaw Ankled, Iran War Ceasefire, Israel's Influence
2026-04-10 spoken.md · speaker-labeled ▶ watch ← E267 all episodes E269 →
0
ideas born
5
ideas moved
7
captures · 3 voices
2
dissenting
+36.0
conviction added
-47.3
decay · 70 silent
Every number here comes from the conviction ledger — the same one the pool ranks on. Decay is what the 70 ideas nobody mentioned gave up this week; it applies only when an episode is processed.
Tier crossings
conviction thresholds crossed by this episode — 65 / 45 / 15 · ideas born here show where they landed. why 65 isn't always green
▼ -9.6
🤖 The $3-4T software industrial complex contracts in 2026
green threshold
→
watch
74.3 → 64.7
▲ +14.7
🤖 Frontier AI labs keep durable premium pricing (not commoditized)
ember
→
watch
33.6 → 48.4
Kill dates that landed since E267
1 hit · 0 partial · 1 miss — windows that closed after 2026-04-03 and up to 2026-04-10, auto-scored against price data and never hand-set. result · return · vs SPY
| idea | result | return | vs SPY | scored |
|---|---|---|---|---|
| 🏛️ China retaliates by openly abandoning IP protection | WRONG | -5.2% | -36.7 | 2026-04-05 |
| 📈 Newsmax's post-IPO meme premium reverts | RIGHT | +88.0% | +56.4 | 2026-04-05 |
Who moved the board
each voice's push on conviction this episode, in conviction points — supports and opposes · share = % of this episode's movement
What got argued (5 ideas)
ordered by how hard each idea moved · quotes are verbatim from the transcript, timestamps deep-link into the episode
▲ MSFT
🤖 Frontier AI labs keep durable premium pricing (not commoditized)
ember 38
POD DISAGREES
▲ +14.7
33.6 → 48.4
I believe open source is going to win the day on the large language models and take 90% of the token usage, and I think the entire frontier model space could be undercut by open source... SLMs, the smaller language models that are verticalized now, that will run on desktops... that is their biggest competitive threat, and I hope it happens.
these revenue ramps are happening while the world is already using open source. This is not frontier only. ... The TAM of intelligence is dramatically larger than any TAM we've ever seen
the drawdown in this period over Iran was only down about five to seven percent on S&P and Nasdaq. ... if we land the plane on these two things, I think it's off to the races in the market. ... heading into America 250, July 4th, the market could really take off.
▼ IGV
🤖 The $3-4T software industrial complex contracts in 2026
ember 34
POD DISAGREES
▼ -9.6
74.3 → 64.7
The market is probably being a little too pessimistic with respect to at least some of these software companies... software is going to be a lot cheaper and easier to generate, but I'm not sure that was the competitive advantage of a lot of these companies. So there's probably a little bit of the baby being thrown out with the bathwater... there probably are some value buys in enterprise software.
▲ GOOGL
🤖 Anthropic IPO (~Oct 2026) is attractively priced at $1.5-2T
watch 61
POD DISAGREES
▲ +9.5
47.2 → 56.7
We've got billions in both companies at this point... They added 4 billion of revenue in January, 7 billion in February, 10 or 11 billion in March [annualized]... the largest revenue explosion in the history of technology... I would not be shocked if you see Anthropic exiting this year at $80 to $100 billion in revenue... I'm a buyer of the shares today.
Episode digest
written during extraction and stored in data/extractions/ep268.json — the auditable source of truth, including everything market-adjacent that did not earn a capture
Friedberg out, Brad in. Anthropic withholds its new Mythos model as too dangerous (autonomously chains 3-5 vulnerabilities; found 27-year-old OpenBSD and 16-year-old FFmpeg bugs), launching Project Glasswing — a 100-day, 40-company hardening sprint: Brad credits market-led self-regulation, Sacks grants this one is legit despite Anthropic's documented fear-marketing pattern (the blackmail study), Chamath calls it 'mostly theater' citing the GPT-2 replay ('you'd have to shut down the internet for five years to patch them all') while separately crowning Anthropic 'shooting the lights out... Steph Curry going bananas.' The $30B run-rate segment is the ledger's biggest lab moment: Brad discloses BILLIONS invested across Anthropic AND OpenAI, projects Anthropic exiting 2026 at $80-100B revenue, says gross margins have swung hugely positive ('accidental profitability' possible) — while Chamath keeps the honesty counterweight (gross-vs-net RevRec confusion, 'how gross margin negative is this? We don't know', enterprise-grade AI coding 'is still s**t', 50 years of tech debt takes decades — softening but not reversing his stance; no clean capture). OpenClaw gets ankled: Anthropic cuts subscription access, clones features, releases its own agent harness — Sacks sketches the bundling/antitrust argument (Anthropic >50% of coding tokens = dominant share); Jason's response is the strongest frontier-ai-premium oppose yet (open source takes 90% of tokens) plus a Ridges-AI reinforcement of his TAO idea; Sacks takes the other side with the coding-flywheel monopolization case, and Brad rebuts commoditization with the frontier-plus-open-source data. Sacks flags value buys in beaten-down enterprise software (tempering oppose on software-industrial-complex). Iran: two-week ceasefire, JD Vance to Islamabad; Brad's Trump-doctrine victory lap ends in an 'off to the races... market could really take off' call attached to us-boom-2026. Labels: clean; Brad properly labeled throughout.