ax@ax-radar:~/curated $ grep -l 'curated=true' sources/
41 srcsignal 72%cycle 04:32

curated · 2026-06-14

▸ 4 items · updated 3m ago
browse by dayclear filter ✕
May 2026
MTWTFSS
1234567368199211010113212281335143815281617179182419472026213222362372462572625272928342936308316
June 2026
MTWTFSS
13823533242253261078828943102711291217138144152161718192021222324252627282930
2026-06-14 · Sun
16:15
17h ago
STILL DEVELOPING · 1dFEATUREDAI HOT (Curated Pool)· aihot-apiZH16:15 · 06·14
→Gary Marcus calls White House AI regulation decision biased toward OpenAI and Amazon, urges independent agency
Gary Marcus argues the White House's Friday action against Anthropic reeks of favoritism. The decision helped OpenAI and Amazon—OpenAI president Greg Brockman is a major Trump donor, and Jared Kushner's brother Josh is a big OpenAI investor. Defense Secretary Pete Hegseth publicly boasted about kicking Anthropic out of the Pentagon three months ago, making the move feel personal. Marcus acknowledges Anthropic overhyped its Mythos model, but says the government gave the company less than 24 hours to respond, relying on an Amazon-triggered report. David Sacks' follow-up statement was desperately vague on what the actual risk was and whether it was unique to Fable/Mythos. The fallout: global customers will rush toward sovereign AI from Europe, Canada, or China rather than bet on US labs that can be shut down without warning. Marcus cites Anthropic's own statement and Cato Institute's Kevin Frazier, both demanding transparent, fair, evidence-driven process. Congressman Ro Khanna proposed an independent agency—Marcus calls that the only way forward.
#Gary Marcus#OpenAI#Anthropic
why featured
Gary Marcus directly names potential conflicts of interest in the White House's ban on Anthropic, providing a concrete chain of personal and financial connections. The piece comes from an influential AI commentator and touches the hottest current AI regulation controversy. The...
editor take
Gary Marcus argues the White House's rushed ban on Anthropic reeks of favoritism toward OpenAI and Amazon.
sharp
Marcus isn't defending Mythos—he says Anthropic overhyped it. His focus is the process: Amazon triggered the review, the White House gave Anthropic less than 24 hours to respond, and David Sacks' follow-up statement was vague on what the actual risk was. Defense Secretary Hegseth publicly boasted about kicking Anthropic out three months ago, which makes the whole thing feel personal. Marcus' real warning is about second-order effects: global customers will now rush toward sovereign AI from Europe, Canada, or China rather than bet on US labs that can be shut down without warning. He cites Cato's Kevin Frazier and Congressman Ro Khanna's call for an independent agency—more of a political signal than a near-term fix.
HKR breakdown
hook ✓knowledge ✓resonance ✓
→ open source
78
SCORE
H1·K1·R1
15:33
18h ago
STILL DEVELOPING · 1dFEATUREDAI HOT (Curated Pool)· aihot-apiZH15:33 · 06·14
→Satya Nadella: A frontier without an ecosystem is unstable
Nadella argues enterprises need both human capital (knowledge, judgment, relationships) and token capital (in-house AI capability). He warns that if a few models capture all value, it repeats the hollowing-out of globalization. The fix: every firm builds its own learning loop—swap base models without losing expert knowledge, and use private evaluation plus RL on real internal trajectories.
#Benchmarking#Microsoft#Satya Nadella
why featured
Nadella himself laying out an enterprise AI strategy with two concrete concepts — token capital and private learning loops — is real signal, not fluff. Held at 78 rather than 85 because we only have the tweet and summary; the full argument isn't spelled out yet.
editor take
Nadella frames in-house AI learning loops as the antidote to value capture by a few frontier models, but skips the how-to.
sharp
This one's worth opening because Nadella rarely spells out enterprise AI strategy this bluntly. His core worry: if a few frontier models capture all the value, companies get hollowed out the way globalization did to some regions. His fix is every firm building its own learning loop—swap base models freely, but keep expert knowledge, private evals, and internal operational data in-house, using RL to make models smarter about your specific business over time. The framework sounds reasonable, but the post doesn't unpack the hard parts: how to build private evals, how to label internal trajectories, how to define the RL reward. Those are where implementation actually stalls. Also, "token capital" feels like a branding wrapper for "in-house AI capability as an asset." I'd read this as Microsoft's strategic pitch to the enterprise market—it wants customers to believe building their own loop on Azure is safer and more durable than just buying API calls.
HKR breakdown
hook ✓knowledge ✓resonance ✓
→ open source
78
SCORE
H1·K1·R1
14:27
19h ago
STILL DEVELOPING · 1dAI HOT (Curated Pool)· aihot-apiZH14:27 · 06·14
→Qiaomu Novel Generator Skill open-sourced: one sentence to start AI novel writing
The open-source Qiaomu Novel Generator Skill lets users say 'I want to write a novel' or specify a style, and AI auto-generates plot outline, character settings, hooks, classic scenes, character desires, conflict escalation, and ending. After discussion and confirmation with AI, it produces a complete novel with low AI-taint. Install via: npx skills add joeseesun/qiaomu-novel-generator. GitHub repo in comments. The post doesn't spell out model support, generation speed, or output length limits.
#Qiaomu Novel Generator#joeseesun#Open source
why featured
An open-source novel generation Skill with a full feature list but no quality benchmarks or real output samples. Hooks with 'AI writes a novel for you,' but lacks K (no quality data, length limits, or Chinese support details) and R is niche. Interesting but thin — fits the 'al...
editor take
Open-source novel generator Skill from prompt to full story, but the post doesn't say which models it supports or output limits.
HKR breakdown
hook ✓knowledge —resonance —
→ open source
55
SCORE
H1·K0·R0
02:56
1d ago
AI HOT (Curated Pool)· aihot-apiZH02:56 · 06·14
→Anthropic CEO Dario Amodei drops bombshells: Claude used by US military in Iran war, 150 girls killed
Anthropic CEO Dario Amodei dropped a series of bombshells ahead of the IPO. He revealed internal model Mythos has thousands of vulnerabilities that could hack banks and steal state secrets. Claude has been used by the US military in the Iran war, involving the killing of 150 girls at a school. He predicts AI will eliminate half of entry-level white-collar jobs in 1-5 years and puts civilization collapse probability at 10%-25%. He left OpenAI due to broken trust and pushed back against Jensen Huang's doomsday marketing accusations. The post doesn't spell out what Mythos is, the specific vulnerabilities, or the IPO timeline.
#Anthropic#Dario Amodei#Claude
why featured
Zero-sourcing content triggers hard exclusion rule #6. The summary explicitly states 'the body does not disclose specific vulnerability details.' No data, no sources, no named examples. Importance capped at 39, tier=excluded.
editor take
Anthropic CEO drops IPO bombshells: internal model Mythos has thousands of vulnerabilities, Claude used by US military in Iran war—but no details on the flaws or IPO timeline.
HKR breakdown
hook ✓knowledge —resonance —
→ open source
39
SCORE
H1·K0·R0

more

✕

feeds

hot eventsdaily columnall postspodcastscuratedX monitorsavedsourcesagent access

admin

usagesystemcurationiterationsusers