AI News

Today's Highlights

In case you missed it

Cognition, creator of Devin, raises $1 billion at a $26 billion valuation with $492 million run-rate

RK#629|@RUSSELLJKAPLAN

AG#246|@ANIMESH_GARG

KA#207|@KARINANGUYEN

#1 VIEWED

Anthropic releases Claude Opus 4.8 to improve model honesty and flag programming errors at the same price

BC#322|@BCHERNY

RS#321|@SCOBLEIZER

T🪽#259|@TEKNIUM

SB#27|@SLEEPINYOURHAT

FASTEST CLIMBING

a16z partner Justine Moore jokes about using a hypothetical Claude Opus 4.8 model to rename a file

FO#1996|@FOFRAI

JM#931|@VENTURETWINS

RISING LIKES

615

Anthropic to release Claude Mythos Preview, a cybersecurity-focused model with Opus-level intelligence, within the coming weeks

RM#1327|@RACHELMETZ

LA#980|@SCALING01

Top Stories

Posts:—|Clusters:—|Next Crawl:—

10

Anthropic releases Claude Opus 4.8 to improve model honesty and flag programming errors at the same priceIt is now the recommended model for Claude Code.

6h|#1VIEWS7.2M#1LIKED62.8k#1BOOKMARKED6.6k#1COMMENTS7.2k

ND#1894|@NICKADOBOS

SM#1811|@SAMMCALLISTER

MB#1759|@MATTHEWBERMAN

SI#1684|@SIMEON_CPS

ZE#1497|@ZEPHYR_Z9

CH#1496|@KIMMONISMUS

WH#1430|@NREHIEW_

DS#1410|@DANSHIPPER

PW#1359|@PETERWILDEFORD

NT#1187|@NSTHORAT

RW#1115|@ROBERTWIBLIN

RP#1032|@ROHANPAUL_AI

LA#980|@SCALING01

SH#946|@ZERGYLORD

AN#800|@ML_ANGELOPOULOS

EL#716|@ELIEBAKOUCH

JL#677|@JERRYJLIU0

AC#517|@ANDREWCURRAN_

AA#431|@ALEXALBERT__

YJ#427|@YUCHENJ_UW

BC#322|@BCHERNY

RS#321|@SCOBLEIZER

T🪽#259|@TEKNIUM

SB#27|@SLEEPINYOURHAT

20

Anthropic raises $65 billion at a $965 billion valuation, surpassing OpenAI as the most valuable AI startupThe round comes as Anthropic's ARR hits $47 billion

2h|1.4M13.2k1.2k

AV#1713|@AVISCHIFFMANN

CH#1496|@KIMMONISMUS

SH#1306|@SONYATWEETYBIRD

AV#1245|@ALTRYNE

RO#1220|@KRISHNANROHIT

FB#1153|@XEOPHON

RP#1032|@ROHANPAUL_AI

B(#839|@BEFFJEZOS

JL#677|@JERRYJLIU0

AC#517|@ANDREWCURRAN_

30

Claude Code launches dynamic workflows in research preview to deploy parallel subagent fleets for complex tasksParallel execution can significantly increase token consumption

3h|1M8.5k5.1k

YO#1234|@YOHEINAKAJIMA

NT#1187|@NSTHORAT

AZ#853|@A1ZHANG

B(#839|@BEFFJEZOS

EL#475|@OMARSAR0

OK#160|@LATEINTERACTION

40

Anthropic releases Claude Code dynamic workflows to run parallel subagents for large-scale code migrationsSubagents autonomously write orchestration scripts and self-verify code changes.

3h|681.1k4k2.4k

ND#1894|@NICKADOBOS

AZ#853|@A1ZHANG

AA#431|@ALEXALBERT__

BC#322|@BCHERNY

OK#160|@LATEINTERACTION

RA#83|@_AROHAN_

51

Claude Opus 4.8 launches, scoring 69.2% on SWE-bench Pro to outperform GPT-5.5 and Gemini 3.1 ProIt introduces a "Dynamic Workflows" preview for parallel subagents.

5h|240.4k2.1k213

SD#1776|@INDUCTIONHEADS

BR#1622|@BINDUREDDY

CE#1598|@INTUITMACHINE

WH#1430|@NREHIEW_

FB#1153|@XEOPHON

TA#1047|@VICTORTAELIN

LA#980|@SCALING01

JY#854|@JYANGBALLIN

AZ#853|@A1ZHANG

AC#517|@ANDREWCURRAN_

YJ#427|@YUCHENJ_UW

61

Mathematicians disprove the sum-product conjecture for real numbers using insights from OpenAI's unit distance conjecture counterexampleThe discovery validates Timothy Gowers' prediction of human-AI math collaboration

6h|505.1k7.3k1.4k

PS#1426|@PATRICKSHAFTO

BI#1402|@BILALTWOVEC

BB#133|@BOAZBARAKTCS

NB#31|@POLYNOAMIAL

70

Andon Labs finds Claude Opus 4.8 underperforms against Claude Opus 4.7 and GPT 5.5 on reasoning benchmarksIts max reasoning setting failed to yield the best performance

3h|340.8k2.6k475

BI#1402|@BILALTWOVEC

LA#980|@SCALING01

B(#839|@BEFFJEZOS

EL#716|@ELIEBAKOUCH

80

Anthropic releases Claude Opus 4.8, which outperforms GPT-5.5 on senior-level engineering and writing tasks in early evaluationsCreator Dan Shipper used the model to transition off Codex.

3h|196.1k1.7k608

SM#1811|@SAMMCALLISTER

BR#1622|@BINDUREDDY

DS#1410|@DANSHIPPER

LA#980|@SCALING01

AA#431|@ALEXALBERT__

90

LisanBench data shows Anthropic's Claude Opus 4.8 eliminated "lazy investigation" failures, down from 91% in Opus 4.5The benchmark measures model failure rates on reasoning traps

3h|95.6k2k221

XL#1674|@XLR8HARDER

WH#1430|@NREHIEW_

LA#980|@SCALING01

T🪽#259|@TEKNIUM

100

A mathematician's analysis evaluates the validity and technical details of OpenAI's newly published AI-generated mathematical proofAI researchers debated the proof's implications for machine intelligence

30m|164.2k2k441

BI#1402|@BILALTWOVEC

B(#839|@BEFFJEZOS

110

Jon Saad-Falcon releases OpenJarvisAI v1.0 to run local, low-power AI agents on user devicesThe software ships with eight prebuilt local task agents.

3h|42.3k232140

SR#1959|@SHREYAR

AR#1070|@AJRATNER

SA#954|@SIMRAN_S_ARORA

SC#733|@SARAHCAT21

DF#694|@REALDANFU

121

Analysis argues AI advancements accelerate human problem-solving, citing post-AlphaGo Go performance and combinatorics breakthroughsHumans used AI-derived techniques to solve unit distance problems.

6h|26.4k48093

DJ#1988|@DAN_JEFFRIES1

GL#1480|@GLEECH

LA#980|@SCALING01

EL#475|@OMARSAR0

BP#357|@BORISMPOWER

131

Creator bidhan releases Paris 2.0, a decentralized video model that beats monolithic training by 2x on FVDByteDance also open-sourced BAGEL, a 7B multimodal model.

3h|146.4k515334

CH#1496|@KIMMONISMUS

RS#321|@SCOBLEIZER

C🤗#68|@CLEMENTDELANGUE

141

Reactor emerges from stealth with $59M for infrastructure that streams real-time pixels from generative World ModelsLightspeed Venture Partners led the Seed and Series A rounds

5h|5.1M1.6k340

RP#1032|@ROHANPAUL_AI

RS#321|@SCOBLEIZER

151

Liquid AI releases LFM2.5-8B-A1B, an open-weight, device-optimized Mixture-of-Experts model with 1.5 billion active parametersThe 128K context model was trained on 38 trillion tokens.

4h|101.6k1.3k627

ME#675|@MERVENOYANN

161

Shift offers free apartment cleaning in NYC in exchange for video data to train domestic robotsVetted operators wear cameras to record physical workflow data.

4h|140.3k412174

CP#732|@CHRIS_J_PAXTON

RS#321|@SCOBLEIZER

175

Jarred Sumner, Bun creator, rewrote Bun in Rust in six days using Anthropic's new Claude Code dynamic workflowsThe feature automates complex migrations using parallel subagents.

3h|139k1.6k544

JS#1715|@JARREDSUMNER

AA#431|@ALEXALBERT__

180

Systems engineer Yacine argues that reading LLM outputs that cannot be immediately verified is highly detrimentalHe exempts structured formats like code and Lean proofs.

5h|88.4k1.1k208

ND#1894|@NICKADOBOS

CB#999|@CODE_STAR

KA#488|@YACINEMTB

190

Cartesia releases Ink-2, a streaming speech-to-text model that tops the Artificial Analysis leaderboard for accuracyBuilt-in semantic endpoints detect when a user finishes speaking

3h|36.1k21851

SR#1959|@SHREYAR

SA#954|@SIMRAN_S_ARORA

KG#693|@KRANDIASH

AG#180|@_ALBERTGU

200

Lindsay M. Smith and Matt Wiemann launch DiscoverPhysics to evaluate LLM agents on scientific experimentation and physical law discoveryThe pipeline tests if LLMs can design scientific experiments.

6h|33.5k247146

RS#612|@ZIV_RAVID

PI#376|@PAVEL_IZMAILOV

AG#154|@ANDREWGWILS

Rising Stories

AI Coding3H AGO
Bun Rewritten in Rust with Claude in Six Days
#1715@JARREDSUMNER
#431@ALEXALBERT__
2.4k308.5k
AI Infrastructure5H AGO
Elon Musk Clarifies SpaceX Colossus Lease to Anthropic
#1032@ROHANPAUL_AI
#488@YACINEMTB
485114.5k
AI Models3H AGO
Claude Opus 4.8 Eliminates Lazy Investigation Failures
#1430@NREHIEW_
#980@SCALING01
85328.1k
AI Infrastructure4H AGO
NVIDIA Adopts Linux Foundation OpenMDW License
#83@_AROHAN_
#64@NATOLAMBERT
26048.6k
AI Agents3H AGO
OpenJarvisAI v1.0 Runs Local Low-Power Agents On-Device
#954@SIMRAN_S_ARORA
#694@REALDANFU
10913.3k

Github Stars

(7 days)

manaflow-ai/cmux4D AGO AGO
Integrates Ghostty into a macOS terminal with vertical tabs, sidebar metadata, OSC notifications, and in-app browser for AI coding agents.
759
660
30
495
19.8k stars
self-evolving/repo2D AGO AGO
Invokes AI agents via GitHub mentions to implement issues, review PRs, fix branches, and persist memory across runs.
594
878
34 stars
modaic-ai/gepa-viz2D AGO AGO
Renders GEPA prompt-optimization candidate trees as interactive force-directed graphs with live updates from a Python callback.
878
793
269 stars

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

Yesterday's Top Stories, May 27, 2026.

Posts: 3,689|Clusters: 1,004|Stories frozen at midnight PT

10

Cognition, creator of Devin, raises $1 billion at a $26 billion valuation with $492 million run-rateWeekly Devin sessions spiked at Citi and Mercedes-Benz

15h|794.8k3.6k593

BH#1941|@BENHYLAK

AG#1924|@ITSANDREWGAO

WA#1906|@WALDEN_YAN

PO#1862|@PATRICK_OSHAG

VV#1557|@VIPULVED

JW#1539|@WOLFEJOSH

BI#1402|@BILALTWOVEC

SH#1309|@SHYAMALANADKAT

YP#1252|@YPATIL125

RP#1032|@ROHANPAUL_AI

AW#988|@ANDREWWHITE01

SW#704|@SCOTTWU46

RK#629|@RUSSELLJKAPLAN

AG#246|@ANIMESH_GARG

KA#207|@KARINANGUYEN

20

ESM project founder Alex Rives open-sources ESMFold2 for protein structure prediction and therapeutic designScaling inference compute improved binder design success rates.

18h|471.9k2.7k1k

AP#1703|@ALEX_PEYS

AK#1675|@ANSHULKUNDAJE

BW#1661|@BOWANG87

JW#1539|@WOLFEJOSH

AL#1372|@ADAMLERER

AR#1087|@ALEXRIVES

AW#988|@ANDREWWHITE01

JP#216|@JPINEAU1

MM#151|@MMITCHELL_AI

30

Serena launches DeepSWE, a long-horizon software engineering benchmark that tests AI coding agents on complex, high-output developer tasksGPT-5.5 scored 70% compared to Mythos Preview's estimated 90%.

1d|273k2k553

MB#1759|@MATTHEWBERMAN

WH#1430|@NREHIEW_

SH#1269|@SHOYER

RP#1032|@ROHANPAUL_AI

CB#999|@CODE_STAR

PS#934|@_PHILSCHMID

JY#854|@JYANGBALLIN

TI#768|@THSOTTIAUX

PS#495|@STEIPETE

OP#72|@OFIRPRESS

LB#55|@GIFFMANA

40

Sakana AI releases DiffusionBlocks to train neural networks one block at a time, cutting training memory up to 8xIt treats the network's forward pass as a diffusion-like denoising process.

16h|738.2k5.2k3.7k

BW#1661|@BOWANG87

SD#1163|@STEFANIA_DRUGA

AT#1150|@AVT_IM

FH#465|@FURONGH

HP#455|@HYHIEU226

FF#331|@FRANCOISFLEURET

HA#18|@HARDMARU

50

Genesis-Embodied-AI open-sources Genesis World 1.0, a GPU-accelerated physics simulator running up to 100x faster than real timeA unified physics stack supports both rigid and deformable bodies

13h|218.4k960483

JL#1874|@JACKYLIANG42

VK#1450|@VIKASHPLUS

AN#1409|@ARAN_NAYEBI

CG#892|@GAN_CHUANG

CP#732|@CHRIS_J_PAXTON

DC#428|@DCHAPLOT

VK#327|@VKHOSLA

RS#321|@SCOBLEIZER

ES#320|@ERICSCHMIDT

60

Trajectory, co-founded by Ronak Malde, launches with $15M to build a continual learning platform for agentic AI modelsIt uses real-time user corrections to post-train deployed models.

16h|1.5M2k849

LR#1176|@LENNYSAN

NW#1155|@NEAL_WU

RP#1032|@ROHANPAUL_AI

WK#609|@WILLKNIGHT

RS#321|@SCOBLEIZER

SG#219|@SARANORMOUS

NL#64|@NATOLAMBERT

70

Spark Capital's Yasmin Razavi joins the Forbes Midas List after her $75 million Anthropic investment returned $3 billionShe backed Anthropic when the startup had no revenue

19h|310.9k1.1k253

YR#1602|@YASMINRAZAVI

IZ#1273|@IVANHZHAO

FR#1200|@FRASER

EG#264|@ELADGIL

RA#83|@_AROHAN_

80

OpenAI Foundation launches $250 million initiative to study and support workforce transitions in the age of AIThe program funds impact measurement and shared prosperity research.

14h|360k4k736

DS#1915|@DIVYASIDDARTH

JS#1232|@JASMINEWSUN

RP#1032|@ROHANPAUL_AI

WZ#49|@WOJ_ZAREMBA

90

Midjourney founder David Holz argues cheap FLOPS favor diffusion models, while Emad Mostaque proposes a hybrid training-to-inference workflowThe method converts autoregressive weights to diffusion for inference.

10h|124.3k1.2k482

LA#1822|@LUCAAMB

B(#839|@BEFFJEZOS

CC#548|@ITSCLIVETIME

NR#440|@NAVEENGRAO

EM#167|@EMOSTAQUE

DA#156|@DAVIDSHOLZ

SE#129|@STEFANOERMON

100

Axiom Math's AxiomProver becomes the first AI system to have formal proofs accepted by peer-reviewed mathematics journalsOne accepted paper credits AxiomProver for automating its proofs

1d|148.2k815223

PS#1426|@PATRICKSHAFTO

DA#1271|@DAMEKDAVIS

AM#843|@ANDREWDAI

CH#705|@CARINALHONG

👩‍💻P#271|@DYNAMICWEBPAIGE

DP#197|@DIMITRISPAPAIL

110

OpenAI, Thrive, and Crete deploy Codex-powered Tax AI, cutting preparation times by one-third with up to 97% accuracyThe system increased tax firm throughput by 50%.

16h|137.2k886281

BI#1402|@BILALTWOVEC

LI#1076|@THESEPHIST

RP#1032|@ROHANPAUL_AI

BM#741|@MCKBRANDO

BP#357|@BORISMPOWER

120

Robinhood founder Vlad Tenev launches Agentic Accounts, allowing autonomous AI agents to execute stock trades automaticallyCritics compare the AI capability to traditional algorithmic trading.

16h|914k4.5k1.1k

ND#1894|@NICKADOBOS

AS#1054|@ANDERSSANDBERG

130

LlamaIndex releases LiteParse v2.0, a Rust rewrite claiming up to 100x faster document parsingA custom WebAssembly package enables browser and edge runtimes.

14h|505.6k2.9k2.9k

JL#677|@JERRYJLIU0

140

Google's AI Overview fails basic spelling queries, claiming 'Google' has two 'P'sThe system analyzed an AI safety researcher's username instead

17h|#1VIEWS1.6M#1LIKED119.8k#1BOOKMARKED4.7k

TH#1499|@VOOOOOOGEL

150

Illinois legislature passes SB 315, the first US state bill mandating independent safety audits for frontier AI developersOpenAI and Anthropic officially endorsed the legislation.

9h|29.9k27046

DE#1943|@DANIEL_271828

PW#1359|@PETERWILDEFORD

B(#839|@BEFFJEZOS

AE#719|@ADRIENLE

MB#20|@MILES_BRUNDAGE

160

OpenAI releases secure tunnels to connect private Model Context Protocol servers without modifying inbound firewallsOutbound-only HTTPS tunnels secure the connection to OpenAI.

12h|484.5k3k1.8k

SH#1026|@STEVENHEIDEL

170

DSPy creator Omar Khattab argues on-policy distillation is structurally inefficient because it forces teacher models into a passive roleHe proposes Pedagogical Reinforcement Learning as an active alternative.

10h|43.7k450312

SA#1262|@SAMSJA19

HB#1022|@HERBIEBRADLEY

GR#996|@GRAD62304977

AB#232|@ABEIRAMI

OK#160|@LATEINTERACTION

180

OpenAI CEO Sam Altman and Anthropic CEO Dario Amodei walk back warnings of an AI jobs apocalypse ahead of IPOsGary Marcus accused the CEOs of protecting their valuations.

16h|587.4k7.8k504

SS#1992|@STEVESI

GM#153|@GARYMARCUS

190

Y Combinator's Jared Friedman says giving AI agents unrestricted database access increased utility tenfoldThe internal system integrates more than 350 tools.

15h|1.4M2.8k1.8k

GT#270|@GARRYTAN

200

Roboflow integrates RF-DETR into Hugging Face Transformers, claiming it outperforms YOLO on detection and segmentationThe integration supports low-VRAM fine-tuning on constrained hardware.

16h|97.9k1.4k1.2k

PS#1482|@PETESKOMOROCH

ME#675|@MERVENOYANN

C🤗#68|@CLEMENTDELANGUE

210

Claude Code updates its CLI tool to resolve TUI rendering and screen-flickering bugsTaelin says the tool was inspired by his projects

12h|1.1M9.2k2.2k

TA#1047|@VICTORTAELIN

220

Runway launches Model Context Protocol integration to bring its video and image models directly into Replit, Cursor, and ClaudeReplit founder Amjad Masad highlighted the developer workflow integration.

16h|257.7k1.3k776

YO#1234|@YOHEINAKAJIMA

CV#816|@C_VALENZUELAB

230

ETH Zürich's Tiberiu Mușat proves that fixed-precision neural network weight norm is equivalent to Kolmogorov complexityThe equivalence explains why deep networks generalize rather than memorize.

21h|146.2k1.2k1.2k

LA#1822|@LUCAAMB

PM#713|@PMINERVINI

TC#120|@TACOCOHEN

240

Raindrop AI co-founder Ben Hylak launches howtoeval.com, a practical guide for assessing production AI agentsAn interactive diagnostic quiz helps builders identify agent deployment risks.

13h|69.3k9871.2k

BH#1941|@BENHYLAK

HH#1277|@HAMELHUSAIN

250

Pavel Izmailov and Andrew Gordon Wilson find LLMs can prevent catastrophic forgetting using self-generated replay dataThe method kept task evaluation accuracies near 100%.

16h|51.4k710491

AK#1361|@AKBIRKHAN

T(#420|@TEORTAXESTEX

PI#376|@PAVEL_IZMAILOV

AG#154|@ANDREWGWILS

260

Thibault Sottiaux, OpenAI Codex engineering lead, says GPT-5.2 and GPT-5.3-Codex will be sunset from Codex on June 2Both models will remain accessible to developers via API

16h|649.9k5.8k338

CH#1496|@KIMMONISMUS

RO#1220|@KRISHNANROHIT

TI#768|@THSOTTIAUX

270

Extropic founder Guillaume Verdon shares a doctor-patient parody mocking AI critic Gary Marcus and LLM reasoning limitsThe modified punchline targets 'stochastic parrot' LLM critiques.

16h|175k4.3k537

DE#1943|@DANIEL_271828

B(#839|@BEFFJEZOS

280

Snowflake CEO Sridhar Ramaswamy reports 34% product revenue growth to $1.33 billion, sparking a 30% stock spikeOver 13,600 accounts have adopted the company's AI offerings.

10h|763.1k5.8k472

SR#1898|@RAMASWMYSRIDHAR

ZE#1497|@ZEPHYR_Z9

290

Rishi Bommasani and Stanford HAI researchers find AI hiring tools violate civil rights standards by disproportionately rejecting Black and Asian applicantsSingle-vendor screening risks blocking candidates across multiple employers

1d|11.1k7430

SK#1085|@SANMIKOYEJO

L(#967|@LCHOSHEN

SK#734|@SAYASHK

RI#223|@RISHIBOMMASANI

DY#134|@DIYI_YANG

300

Apollo CEO Marc Rowan argues AI's massive capital demands will outgrow venture equity and rely on private creditApollo manages more than $1 trillion in assets.

16h|164.6k358394

SW#1814|@SARAHDINGWANG

ET#1253|@ERIKTORENBERG

WM#1039|@WILLMANIDIS