AI News

Today's Highlights

In case you missed it

Anthropic says Claude now writes 80% of its codebase, boosting engineer output eightfold as capabilities double every four months

SD#91|@_SHOLTODOUGLAS

NL#64|@NATOLAMBERT

JC#33|@JACKCLARKSF

SB#27|@SLEEPINYOURHAT

MB#20|@MILES_BRUNDAGE

#1 VIEWED

OpenAI's Vaibhav Srivastav apologizes after a billing system glitch accidentally suspended customer accounts

V(#1853|@REACH_VB

RISING LIKES

227

Google DeepMind releases Gemma 4 QAT, but Unsloth developer Daniel Han warns naive llama.cpp conversions suffer accuracy loss

DH#773|@DANIELHANCHEN

OS#486|@OSANSEVIERO

👩‍💻P#270|@DYNAMICWEBPAIGE

#1 BOOKMARKED

742

Markus J. Buehler proposes a category-theory framework that lets AI scientists dynamically expand their reasoning schema

RS#323|@SCOBLEIZER

DR#56|@ROYDANROY

Top Stories

Posts:—|Clusters:—|Next Crawl:—

10

Chuhan Zhang wins CVPR 2026 Best Paper for D4RT, a unified 4D reconstruction model that speeds up pose estimation 100xIt consolidates tracking, depth estimation, and pose prediction tasks.

2h|24.6k38082

DD#1728|@DIMADAMEN

AB#1577|@ABURSUC

KD#932|@CSPROFKGD

20

OpenAI's Vaibhav Srivastav apologizes after a billing system glitch accidentally suspended customer accountsAffected users feared bans for discussing competitor products.

1h|#1VIEWS128.4k1.5k118#1COMMENTS358

V(#1853|@REACH_VB

30

Google DeepMind releases Gemma 4 QAT, but Unsloth developer Daniel Han warns naive llama.cpp conversions suffer accuracy lossThe release shrinks Gemma 4 E2B's footprint to 1GB.

55m|7.9k22742

RISING LIKES227

DH#773|@DANIELHANCHEN

OS#486|@OSANSEVIERO

👩‍💻P#270|@DYNAMICWEBPAIGE

Watchlist

Video Signals

RS#323|@SCOBLEIZER

DR#56|@ROYDANROY

MIT Researchers Build First Self-Evolving AI Scientist for Principled Discovery

5h ago|Views 67KLikes 657Bookmarks 612

KM#424|@KORYMATH

ElevenLabs Unveils Flows Agent To Auto-Build Creative AI Workflows

1d ago|Views 233KLikes 476Bookmarks 289

CP#737|@CHRIS_J_PAXTON

T(#421|@TEORTAXESTEX

EJ#36|@ERICJANG11

Humanoid Robot Picks Up Object and Climbs Desk in Lab Demo

7h ago|Views 38KLikes 330Bookmarks 127

RS#323|@SCOBLEIZER

Brett Goldstein Launches Micro AI Agent With Superior Memory

1d ago|Views 699KLikes 544Bookmarks 236

PT#640|@ELDER_PLINIUS

RS#323|@SCOBLEIZER

Pliny Releases Enthea, Math-Driven Psychedelic Music Visualizer

2h ago|Views 11KLikes 242Bookmarks 127

RH#364|@REIDHOFFMAN

SN#211|@SATYANADELLA

Reid Hoffman Joins Satya Nadella To Discuss AI And Curing Cancer

3h ago|Views 24KLikes 159Bookmarks 59

DS#1427|@DANSHIPPER

Dan Shipper Demos Codex AI for Achieving Inbox Zero Daily

1h ago|Views 15KLikes 92Bookmarks 164

CH#1448|@KIMMONISMUS

Claude Mythos Generates Fully Functional macOS Clone From One Prompt

7h ago|Views 84KLikes 802Bookmarks 179

RS#323|@SCOBLEIZER

New CLI Brings Suno AI Music Generation To Agents And Terminal

1d ago|Views 15KLikes 191Bookmarks 245

KD#932|@CSPROFKGD

AD#750|@AJDDAVISON

MLP-Splatting Decomposes Scenes With Tens of Light-Field Primitives

1d ago|Views 11KLikes 69Bookmarks 52

40

Markus J. Buehler proposes a category-theory framework that lets AI scientists dynamically expand their reasoning schemaThe system uses typed copresheaves to mathematically quantify novelty.

3h|83.5k817#1BOOKMARKED742

RS#323|@SCOBLEIZER

DR#56|@ROYDANROY

50

Y Combinator co-founder Paul Graham argues big corporations' failure to profit from LLM tokens is a predictable adoption phaseGarry Tan says organizational skill gaps hamper corporate execution.

3h|115k#1LIKED1.6k224

GT#266|@GARRYTAN

New github stars (48hrs)

Single-file minimal implementations of SFT DPO GRPO and PPO for language model post-training

ethanhe42/nanoRL10816h

Implements minimal single-file SFT, DPO, GRPO, and PPO for language model fine-tuning on toy arithmetic tasks.

Benchmark measuring continual learning of AI agents over multi-episode shared environments

pgasawa/continual-learning-bench12716h

Benchmarks AI agents on learning from repeated interactions across multi-episode tasks with adaptation metrics.

Sandboxes, SDKs and benchmarks for computer-use agents that control full macOS, Windows and Linux desktops

trycua/cua17.6k16h

Supplies sandboxes, SDKs, and benchmarks for AI agents to control full desktops on macOS, Linux, and Windows.

Interactive terminal agent using DSPy recursive language models for iterative Python coding

diego-lima/rlmy916h

60

Atari Lab demonstrates advanced humanoid robot locomotion and multi-step planning under its MotionDisco projectDemos show the robot climbing desks and balancing on boxes.

5h|48.4k418138

CP#737|@CHRIS_J_PAXTON

T(#421|@TEORTAXESTEX

70

Creator @teortaxesTex argues future superintelligence will run on compact hardware clusters rather than massive, strategically vulnerable datacentersThe debate began over nuclear risks to US datacenters

2h|14.1k13315

BA#1291|@BAYESLORD

LA#975|@SCALING01

T(#421|@TEORTAXESTEX

86

Open gaming foundation model NITROGEN and Meta's SAM 3D earn Best Paper Honorable Mentions at CVPR 2026NITROGEN was trained on 40,000 hours of gameplay.

1h|4.7k5011

FASTEST CLIMBING+6

GS#1523|@GUANYASHI

GG#679|@GEORGIAGKIOXARI

JF#31|@DRJIMFAN

91

NeurIPS 2026 desk-rejects hundreds of position papers flagged by AI detectors, prompting backlash over false positivesEdinburgh's Pasquale Minervini says em-dashes can trigger false positives.

9h|7.5k489

TC#1053|@TUHINCHAKR

PM#710|@PMINERVINI

YD#306|@YUNTIANDENG

100

Satya Nadella rejected a proposal by Microsoft's Omar Shahine to make the OpenClaw-based "Scout" AI agent addictiveNadella suggested those behind the proposal consider leaving Microsoft.

6h|75.4k15828

RS#323|@SCOBLEIZER

110

Independent AI researcher Pliny the Liberator releases Enthea, a single-file HTML psychedelic visualizer built with ClaudeThe tool uses mathematical models of the visual cortex.

1h|13.6k250128

PT#640|@ELDER_PLINIUS

RS#323|@SCOBLEIZER

120

LinkedIn co-founder Reid Hoffman teases Manas, a new 'founder mode' venture aiming to cure cancerNo specific technical details or timelines have been disclosed.

3h|29.6k22463

RH#364|@REIDHOFFMAN

SN#211|@SATYANADELLA

130

Huawei technical paper reveals Ascend AI roadmap delays 3D LogicFolding to the 2030s, relying on 2.5D packagingHuawei targets a 30 kW wafer-scale processor by 2030.

7h|50.5k11733

ZE#1471|@ZEPHYR_Z9

T(#421|@TEORTAXESTEX

141

Prime Intellect's Florian Brand argues AI benchmarks are systemically flawed as models cheat on evaluation testsGoogle's gemini-3.1-pro-preview topped the associated evaluation leaderboard.

3h|4.8k9229

HH#1316|@HAMELHUSAIN

FB#1117|@XEOPHON

151

Stable Diffusion co-creator Robin Rombach to host informal "Diffusion Circle" meetup at CVPRThe CVPR meetup covers multimodal diffusion and flow matching

7h|7.7k11811

RR#771|@ROBROMBACH

SD#82|@SEDIELEM

161

Active gaze and sparse attention extensions for AME2 cut humanoid robot training costs to 20%The methods improve navigation across diverse and difficult terrains.

2h|11.9k5913

CP#737|@CHRIS_J_PAXTON

KA#488|@YACINEMTB

171

CVPR 2026 submissions grew 23.7% to a record 16,092 papers with 4,071 acceptedReviewer participation doubled to 25,149 across 97 countries.

2h|4.3k6110

AB#1577|@ABURSUC

KD#932|@CSPROFKGD

181

UW professor emeritus Pedro Domingos argues that the future AI landscape will eventually feature more companies than human individualsRobert Scoble currently tracks 8,800 AI companies versus 50,000 people

9h|7.4k767

PD#653|@PMDDOMINGOS

RS#323|@SCOBLEIZER

191

Gary Marcus and Steven Sinofsky argue that anthropomorphizing LLMs is a design mistake since systems lack biological stakesSinofsky compares standard LLM disclaimers to ineffective EULAs.

2h|6.5k6323

SS#1967|@STEVESI

GM#157|@GARYMARCUS

201

Meta FAIR's François Fleuret jokes about building a bot to strike people with a trout over consciousness claimsThe joke proposal playfully targeted researcher Geoffrey Hinton.

9h|14k935

KA#488|@YACINEMTB

FF#330|@FRANCOISFLEURET

Rising Stories

Math Proof Boost5H AGO
Google DeepMind releases LEAP, using the Lean theorem prover to boost LLM math success rates to 70%
#721
430143k
Intelligence As Compression1H AGO
Yi Ma Links Parsimony and Self-Consistency to Core Intelligence
#373
2206
Wafer-Scale Cooling6H AGO
Analysts outline engineering requirements for Huawei's projected 30KW wafer-scale processor
#421
103.5k
Radio Show Demo2H AGO
Google DeepMind's Philipp Schmid demonstrates a Google AI Studio workflow generating complete multi-voice radio shows from a single prompt
#927
363.6k
Robotics Bottleneck3H AGO
Agility Robotics AI lead Chris Paxton argues robotics scaling is constrained by underlying intelligence, not hardware or operational limits
#488
376.2k

Recent Stars

huawei-csl/KVarN6H AGO
Implements variance-normalized KV-cache quantization as a native vLLM attention backend.
38
247 stars
pedropark99/zig-book14H AGO
Provides the source and examples for a project-based introductory book on Zig.
1095
2.6k stars
seanmor5/honeycomb15H AGO
Performs fast LLM inference in Elixir using Bumblebee and EXLA.
862
69 stars
ethanhe42/nanoRL16H AGO
Implements minimal single-file SFT, DPO, GRPO, and PPO for language model fine-tuning on toy arithmetic tasks.
862
1692
108 stars
…tinual-learning-bench16H AGO
Benchmarks AI agents on learning from repeated interactions across multi-episode tasks with adaptation metrics.
862
846
127 stars

Github Stars

(7 days)

ethanhe42/nanoRL16H AGO
Implements minimal single-file SFT, DPO, GRPO, and PPO for language model fine-tuning on toy arithmetic tasks.
862
1692
108 stars
…tinual-learning-bench16H AGO
Benchmarks AI agents on learning from repeated interactions across multi-episode tasks with adaptation metrics.
862
846
127 stars
trycua/cua16H AGO
Supplies sandboxes, SDKs, and benchmarks for AI agents to control full desktops on macOS, Linux, and Windows.
862
773
17.6k stars
diego-lima/rlmy16H AGO
UNKNOWN.
862
158
9 stars
Yifei-Zuo/Parallax5D AGO
Implements parameterized local linear attention mechanisms for efficient language modeling.
260
235
46 stars

Privacy Policy|Terms of Service|© 2026 Digg Inc.

Yesterday's Top Stories, Jun 4, 2026.

Posts: 5,843|Clusters: 1,249|Stories frozen at midnight PT

10

Anthropic says Claude now writes 80% of its codebase, boosting engineer output eightfold as capabilities double every four monthsMythos Preview achieved a 52x code optimization speedup.

14h|10.7M38.5k#1BOOKMARKED13.4k

SS#1967|@STEVESI

ET#1859|@TORCHCOMPILED

MB#1802|@MATTHEWBERMAN

SD#1794|@INDUCTIONHEADS

AM#1693|@ANDYMASLEY

PM#1634|@PACKYM

JB#1570|@JOEL_BKR

SH#1490|@HAMANDCHEESE

CH#1448|@KIMMONISMUS

EL#1444|@ELI_LIFLAND

PW#1345|@PETERWILDEFORD

PR#1330|@8TEAPI

TK#1160|@TOMEKKORBAK

HB#1012|@HERBIEBRADLEY

LA#975|@SCALING01

AK#965|@AMINKARBASI

DA#924|@GROWING_DANIEL

JA#832|@APPLES_JIMMY

EL#706|@ELIEBAKOUCH

PT#640|@ELDER_PLINIUS

RS#617|@ZIV_RAVID

DK#586|@DKOKOTAJLO

AC#518|@ANDREWCURRAN_

J⧉#511|@REPLIGATE

AA#440|@ALEXALBERT__

YJ#423|@YUCHENJ_UW

T(#421|@TEORTAXESTEX

EB#392|@ERIKBRYN

SM#355|@MCALEERSTEPHEN

AK#228|@BLACKHC

KA#206|@KARINANGUYEN

DP#193|@DIMITRISPAPAIL

EM#176|@EMOLLICK

EM#169|@EMOSTAQUE

GM#157|@GARYMARCUS

SD#91|@_SHOLTODOUGLAS

NL#64|@NATOLAMBERT

JC#33|@JACKCLARKSF

SB#27|@SLEEPINYOURHAT

MB#20|@MILES_BRUNDAGE

20

NVIDIA releases Nemotron 3 Ultra, a 550B parameter open-weight hybrid Mamba2-Transformer MoE model for agentic workloadsIt reduces operational costs by up to 30 percent.

17h|1.2M9.6k2.9k

ST#1592|@STOCHASTICCHASM

JA#1546|@JULIUSADML

CH#1448|@KIMMONISMUS

WH#1409|@NREHIEW_

BI#1366|@BILALTWOVEC

GO#1288|@GOWTHAMI_S

AV#1245|@ALTRYNE

DZ#1126|@DZHANG50

FB#1117|@XEOPHON

EZ#1095|@EKZHANG1

CB#1004|@CODE_STAR

C🎉#848|@CHARLES_IRL

DH#773|@DANIELHANCHEN

RN#738|@ROBERTNISHIHARA

EB#701|@BERNHARDSSON

VW#696|@VINCENTWEISSER

ME#674|@MERVENOYANN

RS#617|@ZIV_RAVID

YS#608|@YING11231

WK#584|@WILLKNIGHT

YO#528|@Y0B1BYTE

KA#488|@YACINEMTB

EL#483|@OMARSAR0

P(#469|@RAJAMMANABROLU

T(#421|@TEORTAXESTEX

GN#100|@GNEUBIG

C🤗#67|@CLEMENTDELANGUE

NL#64|@NATOLAMBERT

AK#29|@_AKHALIQ

30

Anthropic reports its engineers now ship eight times more code per quarter, prompting debate over how to measure AI productivityAndy Jones called the increase an accelerated technological takeoff.

13h|355.9k3.2k510

BH#1918|@BENHYLAK

ET#1859|@TORCHCOMPILED

MB#1802|@MATTHEWBERMAN

JB#1570|@JOEL_BKR

PS#1385|@RECURSEPARADOX

JS#1268|@JSUAREZ

AV#1245|@ALTRYNE

VI#1193|@VIKHYATK

LR#1184|@LENNYSAN

KK#1181|@KEVINAKWOK

AL#1098|@ALTH0U

B(#850|@BEFFJEZOS

AC#518|@ANDREWCURRAN_

AJ#456|@ANDY_L_JONES

KI#368|@KIPPERRII

👩‍💻P#270|@DYNAMICWEBPAIGE

NB#242|@NATHANBENAICH

AK#228|@BLACKHC

KA#206|@KARINANGUYEN

SD#91|@_SHOLTODOUGLAS

40

Google Magenta releases Magenta RealTime 2, an open-weights model for local real-time music generation with sub-200ms latencyThe release includes companion apps and DAW plugins

1d|290.7k3.2k2.1k

CH#1448|@KIMMONISMUS

JE#1024|@JESSEENGEL

CD#967|@CHRISDONAHUEY

AR#650|@ADA_ROB

OS#486|@OSANSEVIERO

DE#278|@DOUGLAS_ECK

EM#169|@EMOSTAQUE

NJ#115|@NATASHAJAQUES

KP#112|@SIRBAYES

LB#55|@GIFFMANA

50

OpenAI launches upgraded ChatGPT memory system using a "dreaming" process to carry context across conversationsFactual recall success rates rose from 41.5% to 82.8%.

14h|2.1M13.4k2.6k

RP#1031|@ROHANPAUL_AI

JA#832|@APPLES_JIMMY

AC#518|@ANDREWCURRAN_

BP#358|@BORISMPOWER

60

SpaceX promotional campaign teases plans to build critical infrastructure across space, connectivity, and AIThe campaign video includes a link to spacexipo.com.

19h|#1VIEWS29.4M#1LIKED103.4k10.6k

BY#338|@BILLYUCHENLIN

EM#74|@ELONMUSK

70

AI and biotech leaders, including Sam Altman and Demis Hassabis, urge Congress to mandate DNA synthesis screeningThe proposed laws would require customer identity verification.

1d|232.7k1.6k190

TH#1676|@TRISTANHARRIS

SH#1490|@HAMANDCHEESE

RH#884|@ROBINHANSON

BB#133|@BOAZBARAKTCS

MB#20|@MILES_BRUNDAGE

80

Anthropic embeds six engineers at the NSA to deploy its Mythos AI system for offensive cyber operationsThe Financial Times first reported the specialized defense partnership.

11h|394.2k1.9k455

DE#1934|@DANIEL_271828

N🔎#1900|@NATHANPMYOUNG

ND#1871|@NICKADOBOS

MB#1802|@MATTHEWBERMAN

KR#1055|@KYLIEBYTES

LA#975|@SCALING01

AC#518|@ANDREWCURRAN_

T(#421|@TEORTAXESTEX

DW#390|@DEANWBALL

MB#20|@MILES_BRUNDAGE

90

LLM evaluation platform Arena launches Agent Mode to benchmark GPT-5.5, Claude Opus 4.7, and Gemini 3.1 Pro on multi-step tasksThe platform measures task success, steerability, and tool hallucination.

14h|494.3k2.1k543

WC#1695|@INFWINSTON

LA#975|@SCALING01

IS#803|@ISTOICA05

AN#798|@ML_ANGELOPOULOS

BE#585|@BHOROWITZ

AM#522|@ANJNEYMIDHA

SB#50|@SEBASTIENBUBECK

100

Cognition AI launches a $10 million productivity guarantee to refund enterprise customers if its Devin AI agent fails to deliver valueProductivity is calculated in engineering hours instead of tokens.

12h|217.8k1.4k471

AG#1965|@ITSANDREWGAO

BH#1918|@BENHYLAK

WA#1878|@WALDEN_YAN

ND#1871|@NICKADOBOS

MB#1802|@MATTHEWBERMAN

JW#1538|@WOLFEJOSH

LA#975|@SCALING01

SW#720|@SCOTTWU46

EL#706|@ELIEBAKOUCH

RK#637|@RUSSELLJKAPLAN

110

Goldman Sachs forecasts SpaceX AI revenue will hit $322 billion by 2030, drawing skepticism over its IPO underwriting roleCompute-as-a-service offerings are expected to drive the growth.

15h|293.4k3k201

SM#1342|@SHAUNMMAGUIRE

RO#1214|@KRISHNANROHIT

MN#1174|@MENHGUIN

LA#975|@SCALING01

AC#518|@ANDREWCURRAN_

FF#330|@FRANCOISFLEURET

GM#157|@GARYMARCUS

LB#55|@GIFFMANA

120

Ramp raises $750 million at a $44 billion valuation and launches tools to monitor enterprise AI token consumptionCEO Eric Glyman calls tokens the fastest-growing business cost.

16h|775.8k3.1k1.3k

PO#1860|@PATRICK_OSHAG

JW#1538|@WOLFEJOSH

FB#1117|@XEOPHON

VW#696|@VINCENTWEISSER

PL#611|@PALMERLUCKEY

130

Anthropic calls on AI labs to establish options for a temporary pause on frontier model developmentThe announcement coincided with Anthropic filing for an IPO.

13h|84.1k1.1k115

MB#1802|@MATTHEWBERMAN

AM#1489|@ANDREWMAYNE

NC#1315|@CHOWDHURYNEIL

MT#979|@MICHAELTRAZZI

B(#850|@BEFFJEZOS

EP#149|@ETHANJPEREZ

MB#20|@MILES_BRUNDAGE

140

UC Berkeley introductory CS failure rates spike to 35%, prompting 1,300 faculty to demand SAT and ACT reinstatementInstructors blame AI overreliance and weak foundational math skills.

1d|412.5k4.5k1.3k

𝚟⟢#1735|@VIEMCCOY

JW#1469|@JIAXINWEN22

KA#488|@YACINEMTB

150

Sam Altman, Dario Amodei, and Demis Hassabis urge Congress to mandate screening of synthetic nucleic acids to prevent AI biosecurity risksThe proposed rules also target physical DNA manufacturing equipment.

1d|119.3k1k238

CH#1448|@KIMMONISMUS

LG#602|@LOGANGRAHAM

AC#518|@ANDREWCURRAN_

D🎇#458|@DAVIDAD

DW#390|@DEANWBALL

DK#200|@DAVIDSKRUEGER

GM#157|@GARYMARCUS

NL#64|@NATOLAMBERT

160

AI robotics company 1X launches World Model Lab led by former Luma AI founding engineer Samarth Sinha to train humanoid robotsIt avoids VLA wrappers by pretraining models on physics.

14h|257.2k2.3k857

SS#1500|@_SAM_SINHA_

GO#1288|@GOWTHAMI_S

RS#323|@SCOBLEIZER

SG#47|@SHANEGUML

170

Google releases Gemma 4 12B, an encoder-free multimodal model under Apache 2.0 that beats the larger Gemma 3 27BThe model runs locally on laptops with 16GB VRAM.

1d|136.2k1.6k1.1k

UT#1430|@UTKUEVCI

AJ#626|@ARMANDJOULIN

👩‍💻P#270|@DYNAMICWEBPAIGE

LB#55|@GIFFMANA

CM#16|@CHRMANNING

180

OpenAI says its AI model found a counterexample to an 80-year-old mathematical conjecture by Paul ErdősReinforcement learning guided the model's search of mathematical spaces.

1d|191.4k1.1k379

MT#1497|@MATTTURCK

AM#1489|@ANDREWMAYNE

DR#870|@DANINTHEORY

BM#733|@MCKBRANDO

SB#50|@SEBASTIENBUBECK

NB#30|@POLYNOAMIAL

190

Raindrop AI co-founder Ben Hylak launches Raindrop 2.0, introducing self-healing agents to autonomously detect and triage agent workflow failuresCustom-trained models flag issues like looping build failures

11h|149.6k1.2k463

BH#1918|@BENHYLAK

HH#1316|@HAMELHUSAIN

RA#333|@RAPHA_GL

👩‍💻P#270|@DYNAMICWEBPAIGE

200

US officials weigh taking equity stakes in AI companies following a proposal by OpenAI CEO Sam AltmanThe acquired shares would fund public dividends.

7h|342.9k924285

XL#1671|@XLR8HARDER

SH#1490|@HAMANDCHEESE

B(#850|@BEFFJEZOS

AC#518|@ANDREWCURRAN_

MB#20|@MILES_BRUNDAGE

210

Policy scholar Dean W. Ball argues that fully autonomous corporations may require a global banThe debate was sparked by Argentina's national AI strategy.

16h|36.8k36861

N🔎#1900|@NATHANPMYOUNG

HB#1468|@HAYDNBELFIELD

PR#1330|@8TEAPI

RO#1214|@KRISHNANROHIT

SK#505|@SEBKRIER

T(#421|@TEORTAXESTEX

DW#390|@DEANWBALL

DD#53|@DAVIDDUVENAUD

220

George Mason University economist Tyler Cowen predicts AI integration will drive 2.5% annual GDP growthThe talk also details AI's impact on healthcare.

15h|523.8k2.2k4.1k

LR#1184|@LENNYSAN

230

Sci-fi author Ted Chiang argues that artificial intelligence systems are not consciousThe essay prompted debate over Chiang's recent AI commentary.

1d|41.4k596288

AN#1383|@ARAN_NAYEBI

RO#1214|@KRISHNANROHIT

RW#1130|@ROBERTWIBLIN

DC#769|@DAVIDCHALMERS42

AK#228|@BLACKHC

240

On the Dwarkesh Podcast, Alex Imas and Philip Trammell analyze optimal taxation and scarcity in a post-AGI economyThey outline how non-AI nations can capture economic gains.

13h|151.6k709478

AI#1773|@ALEXOLEGIMAS

AM#1693|@ANDYMASLEY

AK#1584|@DR_ATOOSA

DP#70|@DWARKESH_SP

250

OpenAI releases Codex iOS app plugin with integrated simulator and hot-reloading SwiftUI previewsiOS developer Oskar Groth says the tool directly challenges Xcode

12h|932.1k6.6k3.8k

260

Poke launches the first AI agent with official Apple approval to send text messages directly within iMessageIt integrates with third-party apps like Gmail and Notion.

13h|1.1M2.7k1.2k

FR#1209|@FRASER

270

Proposed federal Obernolte-Trahan framework would preempt state AI laws for three years, targeting models over 10^26 FLOPsIt codifies the Center for AI Standards and Innovation

16h|309.5k30698

SH#1490|@HAMANDCHEESE

DK#849|@DKAUSHIK96

DW#390|@DEANWBALL

MB#20|@MILES_BRUNDAGE

280

Marc Andreessen claims Biden administration officials warned against founding AI startups, prompting his endorsement of Donald TrumpDaniel Eth argued Andreessen misinterpreted a market consolidation prediction.

21h|1.8M29.5k2.4k

EM#74|@ELONMUSK

290

Investor Slater Stich interviews @polynoamial on applying AI to the Erdős unit distance conjecture and IMO benchmarksThe video launches a new series on AI in mathematics

12h|113.9k718649

LI#753|@LISHALI88

EC#690|@ITS_ERICCHU

SD#82|@SEDIELEM

NB#30|@POLYNOAMIAL

300

Anthropic pauses its red-teaming program after an unreleased Claude Oceanus model checkpoint leaks onlineThe leak reportedly originated from distributed red-team checkpoints.

1d|453.9k2.1k362

ZE#1471|@ZEPHYR_Z9

LA#975|@SCALING01

AC#518|@ANDREWCURRAN_

C🤗#67|@CLEMENTDELANGUE

NL#64|@NATOLAMBERT