Digg - AI news, before it trends

Highlights

??
IN CASE YOU MISSED IT

OpenAI says an internal general-purpose reasoning model has autonomously produced research-level advances on the Erdős conjecture in discrete geometry by finding new optimal constructions for the planar unit distance problem

Christian SzegedyCS#43|@CHRSZEGEDY
Eric JangEJ#36|@ERICJANG11
Noam BrownNB#31|@POLYNOAMIAL
Miles BrundageMB#20|@MILES_BRUNDAGE
Greg BrockmanGB#19|@GDB
Sam AltmanSA#8|@SAMA
RISING LIKES (5.9k)

Sam Altman, co-founder and CEO of OpenAI, identifies three AGI priority areas of accelerating research, companies, and personal goal achievement while referencing a unit distance result and $2 million credits for YC companies

Beff (e/acc)B(#839|@BEFFJEZOS
Jimmy Apples 🍎/accJA#837|@APPLES_JIMMY
Sam AltmanSA#8|@SAMA
FASTEST CLIMBING (+3)

Antigravity tripled the rate limits for Gemini models across all paid tiers and reset weekly Gemini quotas for users

Varun MohanVM#1144|@_MOHANSOLO
👩‍💻 Paige Bailey👩‍💻P#271|@DYNAMICWEBPAIGE
Logan KilpatrickLK#90|@OFFICIALLOGANK

Top Stories.

Posts: —|Clusters: —|Next crawl: —
10

OpenAI says an internal general-purpose reasoning model has autonomously produced research-level advances on the Erdős conjecture in discrete geometry by finding new optimal constructions for the planar unit distance problem — Features over 100 pages of chain-of-thought and targets a leading mathematics journal.

11h|#1VIEWS12.5M#1LIKED61.4k#1BOOKMARKED14.4k#1COMMENTS4.3k
kalomazeKA#836|@KALOMAZE
You JiachengYJ#826|@YOUJIACHENG
Brendan Dolan-GavittBD#812|@MOYIX
Sherwin WuSW#801|@SHERWINWU
Ekin AkyürekEA#757|@AKYUREKEKIN
Brandon McKinzieBM#741|@MCKBRANDO
Chris PaxtonCP#732|@CHRIS_J_PAXTON
elieEL#716|@ELIEBAKOUCH
Tamay BesirogluTB#699|@TAMAYBES
Aaron RothAR#697|@AAROTH
Bojan TunguzBT#687|@TUNGUZ
Jason PhangJP#672|@ZHANSHENG
Pedro DomingosPD#654|@PMDDOMINGOS
Yo ShavitYS#638|@YONASHAV
Simo RyuSR#604|@CLONEOFSIMO
Patrick HsuPH#597|@PDHSU
heinerHE#572|@HEINRICHKUTTLER
Clive ChanCC#548|@ITSCLIVETIME
Andrew CurranAC#517|@ANDREWCURRAN_
CLSCL#442|@CHENGLEISI
Teortaxes▶️ (DeepSeek 推特🐋铁粉 2023 – ∞)T(#420|@TEORTAXESTEX
Dean W. BallDW#392|@DEANWBALL
Alexander WeiAW#367|@ALEXWEI_
Mo BavarianMB#364|@MOBAV0
Boris PowerBP#357|@BORISMPOWER
Nat McAleeseNM#345|@__NMCA__
François FleuretFF#331|@FRANCOISFLEURET
Andrew Carr 🤸AC#263|@ANDREW_N_CARR
Surya GanguliSG#256|@SURYAGANGULI
will depueWD#254|@WILLDEPUE
Aidan ClarkAC#241|@_AIDAN_CLARK_
Jakub PachockiJP#206|@MERETTM
Michael BronsteinMB#199|@MMBRONSTEIN
Michael NielsenMN#187|@MICHAEL_NIELSEN
Ethan MollickEM#175|@EMOLLICK
EmadEM#167|@EMOSTAQUE
Andrew Gordon WilsonAG#154|@ANDREWGWILS
Gary MarcusGM#153|@GARYMARCUS
Boaz BarakBB#133|@BOAZBARAKTCS
Delip Rao e/σDR#103|@DELIPRAO
Ofir PressOP#72|@OFIRPRESS
Dan RoyDR#56|@ROYDANROY
Lucas Beyer (bl16)LB#55|@GIFFMANA
Sebastien BubeckSB#51|@SEBASTIENBUBECK
Christian SzegedyCS#43|@CHRSZEGEDY
Eric JangEJ#36|@ERICJANG11
Noam BrownNB#31|@POLYNOAMIAL
Miles BrundageMB#20|@MILES_BRUNDAGE
Greg BrockmanGB#19|@GDB
Sam AltmanSA#8|@SAMA
20

Cohere releases Command A+, its most powerful large language model to date, as open-source Apache 2.0 software that runs on two H100 GPUs with 30% lower latency — Cohere co-founder Ivan Zhang highlighted efficiency and accessibility design choices.

15h|666.2k4.6k2k
EthanET#1884|@TORCHCOMPILED
Ivan ZhangIZ#1634|@1VNZH
whWH#1430|@NREHIEW_
Nils ReimersNR#1167|@NILS_REIMERS
Florian BrandFB#1153|@XEOPHON
Cody BlakeneyCB#999|@CODE_STAR
Ben (no treats)B(#982|@ANDERSONBCDEFG
elieEL#716|@ELIEBAKOUCH
Jay AlammarJA#715|@JAYALAMMAR
Pasquale MinerviniPM#713|@PMINERVINI
merveME#675|@MERVENOYANN
Nick FrosstNF#601|@NICKFROSST
kacheKA#488|@YACINEMTB
elvisEL#475|@OMARSAR0
Teortaxes▶️ (DeepSeek 推特🐋铁粉 2023 – ∞)T(#420|@TEORTAXESTEX
Siva ReddySR#370|@SIVAREDDYG
will brownWB#339|@WILLCCBB
Marc G. BellemareMG#258|@MARCGBELLEMARE
Joelle PineauJP#216|@JPINEAU1
Stella Biderman @ ICLRSB#208|@BLANCHEMINERVA
Aidan GomezAG#171|@AIDANGOMEZ
Sebastian RaschkaSR#155|@RASBT
Delip Rao e/σDR#103|@DELIPRAO
clem 🤗C🤗#68|@CLEMENTDELANGUE
AKAK#28|@_AKHALIQ
30

SpaceX files S-1 for Nasdaq IPO under ticker SPCX targeting $1.75 trillion valuation and discloses xAI driving 60 percent of 2025 capital spending — Grok reached 550 million monthly active users with 1 GW of compute deployed.

9h|4.5M34.6k1.4k
Steven SinofskySS#1992|@STEVESI
Nick DobosND#1894|@NICKADOBOS
bilalBI#1402|@BILALTWOVEC
Lisan al GaibLA#980|@SCALING01
Erik BrynjolfssonEB#389|@ERIKBRYN
Elon MuskEM#76|@ELONMUSK
40

Exa raised $250 million in a Series C at a $2.2 billion valuation led by Andreessen Horowitz, reporting 400,000 developers and 5,000 company adopters for its AI agent search platform — Token usage grew 20x for agent-driven queries.

14h|1.2M4.4k2k
andrew gaoAG#1924|@ITSANDREWGAO
Sarah WangSW#1814|@SARAHDINGWANG
mickey friedmanMF#1774|@MICKEYXFRIEDMAN
AviAV#1713|@AVISCHIFFMANN
Will BrykWB#1582|@WILLIAMBRYK
jasonJA#929|@JXNLCO
LishaLI#761|@LISHALI88
Keerthana GopalakrishnanKG#730|@KEERTHANPG
Alexis RossAR#563|@ALEXISJROSS
Marco MascorroMM#557|@MASCOBOT
Garry TanGT#270|@GARRYTAN
swyx🛬 SFOSS#214|@SWYX
Igor BabuschkinIB#117|@IBAB
Paul GrahamPG#78|@PAULG
50

OpenAI announces one of its models identified new constructions for the planar unit distance problem first posed by Paul Erdős in 1946 that outperform square-grid arrangements long assumed optimal — Marks first autonomous AI solution to a longstanding open math problem.

8h|256.7k2k416
Nick DobosND#1894|@NICKADOBOS
thebesTH#1499|@VOOOOOOGEL
Samuel Hammond 🦉SH#1488|@HAMANDCHEESE
Shaun MaguireSM#1333|@SHAUNMMAGUIRE
PrakashPR#1332|@8TEAPI
Alex VolkovAV#1245|@ALTRYNE
Andrea MontanariAM#1088|@ANDREA__M
Miles BrundageMB#20|@MILES_BRUNDAGE
New github stars (48hrs)
aisa-group/InferenceBench211h

Evaluates autonomous AI agents optimizing LLM serving like vLLM in open-ended scenarios with quality and integrity gates.

Rulin ShaoRS1224
Maksym AndriushchenkoMA1066
manaflow-ai/cmux17.6k1d

Integrates Ghostty into a macOS terminal with vertical tabs, sidebar metadata, OSC notifications, and in-app browser for AI coding agents.

Jack MorrisJM203
will brownWB339
sapientinc/HRM-Text5651d

Pretrains 1B HRM text models with hierarchical reasoning, task completion, and low-compute PrefixLM training.

TaelinTA1047
Vincent WeisserVW707
61

SpaceX is hiring world-class engineers and physicists for its SpaceXAI team and states that no prior artificial intelligence experience is required — Elon Musk directed applications to ai_eng@spacex.com with three bullet points.

1h|5.9M27.9k6.1k
TiboTI#768|@THSOTTIAUX
Elon MuskEM#76|@ELONMUSK
71

Anthropic projects $10.9 billion June quarter revenue and its first $559 million operating profit, implying a $44 billion annualized run rate during a funding round likely valuing it above OpenAI. — Limited compute capacity pushes some customers to other providers.

9h|377.2k2.7k440
Timothy B. LeeTB#1556|@BINARYBITS
ZephyrZE#1497|@ZEPHYR_Z9
Chubby♨️CH#1496|@KIMMONISMUS
Lisan al GaibLA#980|@SCALING01
Chris PaxtonCP#732|@CHRIS_J_PAXTON
Andrew CurranAC#517|@ANDREWCURRAN_
81

An AI system solves a widely known open mathematics problem that multiple leading human mathematicians had been unable to resolve after significant effort — Mathematician Timothy Gowers called it one of AI's firsts in the field.

10h|355.9k2k479
Timothy B. LeeTB#1556|@BINARYBITS
Shubhendu TrivediST#1446|@_ONIONESQUE
bilalBI#1402|@BILALTWOVEC
PrakashPR#1332|@8TEAPI
Kevin Weil 🇺🇸KW#451|@KEVINWEIL
Csaba SzepesvariCS#409|@CSABASZEPESVARI
Michael BronsteinMB#199|@MMBRONSTEIN
Thang LuongTL#188|@LMTHANG
91

Jeff Bezos says artificial intelligence will elevate people at work rather than replace jobs, citing assistance in X-ray analysis and programming tasks during a CNBC interview — Bezos linked efficiency gains to potential deflationary effects and cautioned against early regulation.

13h|718.9k8.7k1.4k
Beff (e/acc)B(#839|@BEFFJEZOS
Aaron LevieAL#562|@LEVIE
Andrew CurranAC#517|@ANDREWCURRAN_
100

Intuit is reducing its global workforce by 17 percent, or roughly 3,000 employees, to redirect resources toward AI efforts according to an internal memo reviewed by Reuters — Posts reference Meta cuts and SaaS predictions of 40-50 percent reductions.

17h|873.7k4.2k349
ZephyrZE#1497|@ZEPHYR_Z9
@jason@J#1061|@JASON
Rohan PaulRP#1032|@ROHANPAUL_AI
Andrew CurranAC#517|@ANDREWCURRAN_
110

Terminal-Bench Science extends the original Terminal-Bench benchmark used by Anthropic, OpenAI, and Google DeepMind into scientific domains and opens for over 100 task contributions by August 17, 2026 — Contributors package workflows as RL environments with verification tests.

13h|845.1k579224
Katie KangKK#1481|@KATIE_KANG_
Sanmi KoyejoSK#1085|@SANMIKOYEJO
Alex RatnerAR#1070|@AJRATNER
Lisan al GaibLA#980|@SCALING01
Chenhao TanCT#570|@CHENHAOTAN
Ludwig SchmidtLS#366|@LSCHMIDT3
Alex DimakisAD#172|@ALEXGDIMAKIS
Thomas WolfTW#17|@THOM_WOLF
120

Pirate Wires analysis claims data centers would consume just 8 percent of U.S. golf course water usage by 2030 even after tripling their draw — 400 Arizona golf courses use 42 billion gallons yearly.

13h|798.2k6k1.1k
Mike SolanaMS#1323|@MICSOLANA
Elon MuskEM#76|@ELONMUSK
130

Sam Altman, co-founder and CEO of OpenAI, identifies three AGI priority areas of accelerating research, companies, and personal goal achievement while referencing a unit distance result and $2 million credits for YC companies — Reply asks if personal AGI could build a moon base.

8h|364k5.9k683
RISING LIKES5.9k
Beff (e/acc)B(#839|@BEFFJEZOS
Jimmy Apples 🍎/accJA#837|@APPLES_JIMMY
Sam AltmanSA#8|@SAMA
140

Junyeob Baek and colleagues introduce Generative Recursive reAsoning Models (GRAM) that convert deterministic recursive reasoning into stochastic latent trajectories and report 97.0% accuracy on Sudoku-Extreme with 10 million parameters — The model also scores 52.0% on ARC-AGI-1 and 44.6% on ARC-AGI-2.

17h|158.1k1.4k1.2k
Super DarioSD#1776|@INDUCTIONHEADS
Kyle KastnerKK#1025|@KASTNERKYLE
Mengye RenMR#880|@MENGYER
billy bubbaBB#840|@WGRATHWOHL
150

Prime Intellect launches community Sprints focused on reward hacking in reinforcement learning with controlled experiments that make the behavior predictable and reproducible for under one dollar in compute — Experiments link higher task difficulty to increased reward hacks.

8h|101.3k899427
Florian BrandFB#1153|@XEOPHON
Vincent WeisserVW#707|@VINCENTWEISSER
will brownWB#339|@WILLCCBB
160

Anthropic released a paper outlining US strategies to maintain frontier AI lead over China through advanced chip controls and sustained competition — MIRI's Nate Soares questioned its alignment with safety priorities.

6h|70.2k800131
Nate Soares ⏹️NS#1981|@SO8RES
Oliver HabrykaOH#1914|@OHABRYKA
Nathan is in Berkeley 🔎NI#1902|@NATHANPMYOUNG
Eli LiflandEL#1442|@ELI_LIFLAND
Teortaxes▶️ (DeepSeek 推特🐋铁粉 2023 – ∞)T(#420|@TEORTAXESTEX
Scott AlexanderSA#297|@SLATESTARCODEX
170

The Economist attributes most recent US GDP growth to AI-hardware investment, with that category indexed above 140 by 2026 while overall GDP stays near 100. — Separate note shows AI fixed investment at 5.46 percent of GDP.

14h|113.1k1.3k421
Nathan is in Berkeley 🔎NI#1902|@NATHANPMYOUNG
Alex ImasAI#1777|@ALEXOLEGIMAS
Adrien EcoffetAE#719|@ADRIENLE
180

OpenAI’s o1 model generates original insights to solve complex math problems, marking early recursive self-improvement beyond training data recombination. — Theories of AI as stochastic parrots are now obsolete.

9h|143.5k1.7k401
EthanET#1884|@TORCHCOMPILED
Bojan TunguzBT#687|@TUNGUZ
Andrew CurranAC#517|@ANDREWCURRAN_
190

OpenAI prepares to file confidentially for an initial public offering with Goldman Sachs and Morgan Stanley, potentially as early as Friday for a September 2026 launch — Company is valued at $730 billion after its latest funding round.

14h|280.6k76273
Bojan TunguzBT#687|@TUNGUZ
Andrew CurranAC#517|@ANDREWCURRAN_
roonRO#59|@TSZZL
200

Charts show sharp post-ChatGPT rises in creative and scientific output with weekly Amazon English e-book releases reaching 292,000 and quarterly arXiv submissions hitting 77,621 as AI music uploads overtake non-AI tracks — Federal pro se litigants rose to 17 percent of filings.

16h|175.1k1.1k492
Lenny RachitskyLR#1176|@LENNYSAN
Tuhin ChakrabartyTC#1050|@TUHINCHAKR
Reid HoffmanRH#373|@REIDHOFFMAN

GitHub Stars (7-days)

  1. …ilsonnn/image-blaster4d ago

    Converts an input image into 3D meshes, Gaussian splats, and audio files using Claude skills and external AI APIs.

    Mohit ShridharMS1673
    elvisEL475
    🎭🎭878
    Andrew Carr 🤸AC263
    3.1k stars
  2. …tonomous-speedrunning6d ago

    Archives experiments of AI agents autonomously tuning optimizers, schedules, and hyperparameters to reach target validation loss in fewest steps on a small LM benchmark.

    swyx🛬 SFOSS214
    Vincent WeisserVW707
    samsjaSA1262
    65 stars
  3. …-group/InferenceBench11h ago

    Evaluates autonomous AI agents optimizing LLM serving like vLLM in open-ended scenarios with quality and integrity gates.

    Rulin ShaoRS1224
    Maksym AndriushchenkoMA1066
    2 stars
  4. manaflow-ai/cmux1d ago

    Integrates Ghostty into a macOS terminal with vertical tabs, sidebar metadata, OSC notifications, and in-app browser for AI coding agents.

    Jack MorrisJM203
    will brownWB339
    17.6k stars
  5. sapientinc/HRM-Text1d ago

    Pretrains 1B HRM text models with hierarchical reasoning, task completion, and low-compute PrefixLM training.

    TaelinTA1047
    Vincent WeisserVW707
    565 stars

Recent GitHub Stars

  1. amilich/isometric-city1h ago

    Renders isometric cities with autonomous traffic, pedestrians, zoning and resource simulation on HTML5 Canvas.

    Jack MorrisJM203
    2.1k stars
  2. kageroumado/phosphene4h ago

    Plays user video files as live macOS desktop and lock-screen wallpapers via menu bar app and private WallpaperExtensionKit extension.

    Stephen RollerSR349
    77 stars
  3. …-AI-Alliance/tapestry7h ago

    Builds global frontier LLMs for sovereign AI via consortium training on distributed compute nodes with local data.

    ⿻ Andrew Trask⿻A361
    16 stars
  4. namanbarkiya/inputr9h ago

    Detects image upload constraints on sites then enables crop, create, or draw modes to produce exact matches.

    swyx🛬 SFOSS214
    5 stars
  5. …st-of-Agent-Harnesses9h ago

    Curates and ranks over 100 agent harnesses with weekly scoring, categorized tables, stars, and usage examples.

    🎭🎭878
    16 stars
Privacy Policy|Terms of Service|© 2026 Digg Inc.

Yesterday's Top Stories, May 20, 2026.

Posts: 4,298|Clusters: 1,063|Stories frozen at midnight PT
10

OpenAI says an internal general-purpose reasoning model has autonomously produced research-level advances on the Erdős conjecture in discrete geometry by finding new optimal constructions for the planar unit distance problem — Features over 100 pages of chain-of-thought and targets a leading mathematics journal.

11h|#1VIEWS12.5M61.4k#1BOOKMARKED14.4k
kalomazeKA#836|@KALOMAZE
You JiachengYJ#826|@YOUJIACHENG
Brendan Dolan-GavittBD#812|@MOYIX
Sherwin WuSW#801|@SHERWINWU
Ekin AkyürekEA#757|@AKYUREKEKIN
Brandon McKinzieBM#741|@MCKBRANDO
Chris PaxtonCP#732|@CHRIS_J_PAXTON
elieEL#716|@ELIEBAKOUCH
Tamay BesirogluTB#699|@TAMAYBES
Aaron RothAR#697|@AAROTH
Bojan TunguzBT#687|@TUNGUZ
Jason PhangJP#672|@ZHANSHENG
Pedro DomingosPD#654|@PMDDOMINGOS
Yo ShavitYS#638|@YONASHAV
Simo RyuSR#604|@CLONEOFSIMO
Patrick HsuPH#597|@PDHSU
heinerHE#572|@HEINRICHKUTTLER
Clive ChanCC#548|@ITSCLIVETIME
Andrew CurranAC#517|@ANDREWCURRAN_
CLSCL#442|@CHENGLEISI
Teortaxes▶️ (DeepSeek 推特🐋铁粉 2023 – ∞)T(#420|@TEORTAXESTEX
Dean W. BallDW#392|@DEANWBALL
Alexander WeiAW#367|@ALEXWEI_
Mo BavarianMB#364|@MOBAV0
Boris PowerBP#357|@BORISMPOWER
Nat McAleeseNM#345|@__NMCA__
François FleuretFF#331|@FRANCOISFLEURET
Andrew Carr 🤸AC#263|@ANDREW_N_CARR
Surya GanguliSG#256|@SURYAGANGULI
will depueWD#254|@WILLDEPUE
Aidan ClarkAC#241|@_AIDAN_CLARK_
Jakub PachockiJP#206|@MERETTM
Michael BronsteinMB#199|@MMBRONSTEIN
Michael NielsenMN#187|@MICHAEL_NIELSEN
Ethan MollickEM#175|@EMOLLICK
EmadEM#167|@EMOSTAQUE
Andrew Gordon WilsonAG#154|@ANDREWGWILS
Gary MarcusGM#153|@GARYMARCUS
Boaz BarakBB#133|@BOAZBARAKTCS
Delip Rao e/σDR#103|@DELIPRAO
Ofir PressOP#72|@OFIRPRESS
Dan RoyDR#56|@ROYDANROY
Lucas Beyer (bl16)LB#55|@GIFFMANA
Sebastien BubeckSB#51|@SEBASTIENBUBECK
Christian SzegedyCS#43|@CHRSZEGEDY
Eric JangEJ#36|@ERICJANG11
Noam BrownNB#31|@POLYNOAMIAL
Miles BrundageMB#20|@MILES_BRUNDAGE
Greg BrockmanGB#19|@GDB
Sam AltmanSA#8|@SAMA
20

SpaceX's IPO filing discloses a multi-year compute services agreement with Anthropic under which the AI company will pay $1.25 billion per month through May 2029 — Covers Colossus clusters for Claude inference on NVIDIA GB200 GPUs

14d|11.6M#1LIKED61.5k3.9k
Garrison Lovely is in DCGL#1984|@GARRISONLOVELY
Patrick OShaughnessyPO#1862|@PATRICK_OSHAG
🍓🍓🍓🍓🍓#1711|@IRULETHEWORLDMO
Nitasha TikuNT#1616|@NITASHATIKU
Austen AllredAA#1540|@AUSTEN
ZephyrZE#1497|@ZEPHYR_Z9
Chubby♨️CH#1496|@KIMMONISMUS
bilalBI#1402|@BILALTWOVEC
Shaun MaguireSM#1333|@SHAUNMMAGUIRE
Rachel MetzRM#1327|@RACHELMETZ
signüllSI#1124|@SIGNULLL
Kylie RobisonKR#1046|@KYLIEBYTES
Lisan al GaibLA#980|@SCALING01
Alexander DoriaAD#867|@DORIALEXANDER
Beff (e/acc)B(#839|@BEFFJEZOS
elieEL#716|@ELIEBAKOUCH
Zachary NadoZN#537|@ZACHARYNADO
Andrew CurranAC#517|@ANDREWCURRAN_
Teortaxes▶️ (DeepSeek 推特🐋铁粉 2023 – ∞)T(#420|@TEORTAXESTEX
Tom BrownTB#384|@NOTTOMBROWN
Tanishq Mathew Abraham, Ph.D.TM#359|@ISCIENCELUVR
Robert ScobleRS#321|@SCOBLEIZER
Elon MuskEM#76|@ELONMUSK
Lucas Beyer (bl16)LB#55|@GIFFMANA
30

Cohere releases Command A+, its most powerful large language model to date, as open-source Apache 2.0 software that runs on two H100 GPUs with 30% lower latency — Cohere co-founder Ivan Zhang highlighted efficiency and accessibility design choices.

15h|666.2k4.6k2k
EthanET#1884|@TORCHCOMPILED
Ivan ZhangIZ#1634|@1VNZH
stochasmST#1629|@STOCHASTICCHASM
whWH#1430|@NREHIEW_
Nils ReimersNR#1167|@NILS_REIMERS
Florian BrandFB#1153|@XEOPHON
Ben (no treats)B(#982|@ANDERSONBCDEFG
elieEL#716|@ELIEBAKOUCH
Jay AlammarJA#715|@JAYALAMMAR
Pasquale MinerviniPM#713|@PMINERVINI
merveME#675|@MERVENOYANN
Nick FrosstNF#601|@NICKFROSST
kacheKA#488|@YACINEMTB
elvisEL#475|@OMARSAR0
Teortaxes▶️ (DeepSeek 推特🐋铁粉 2023 – ∞)T(#420|@TEORTAXESTEX
Siva ReddySR#370|@SIVAREDDYG
will brownWB#339|@WILLCCBB
Marc G. BellemareMG#258|@MARCGBELLEMARE
Joelle PineauJP#216|@JPINEAU1
Stella Biderman @ ICLRSB#208|@BLANCHEMINERVA
Aidan GomezAG#171|@AIDANGOMEZ
Sebastian RaschkaSR#155|@RASBT
Delip Rao e/σDR#103|@DELIPRAO
clem 🤗C🤗#68|@CLEMENTDELANGUE
AKAK#28|@_AKHALIQ
40

Google releases Gemini 3.5 Flash with $1.50 per million input token pricing and up to four times faster output than Gemini 3.1 Pro — Real-world usage shows three times higher token consumption than predecessors.

2d|805.3k8.5k714
EthanET#1884|@TORCHCOMPILED
Theo - t3.ggT-#1829|@THEO
Bindu ReddyBR#1622|@BINDUREDDY
Alex VolkovAV#1245|@ALTRYNE
signüllSI#1124|@SIGNULLL
Rohan PaulRP#1032|@ROHANPAUL_AI
Cody BlakeneyCB#999|@CODE_STAR
Ben (no treats)B(#982|@ANDERSONBCDEFG
Lisan al GaibLA#980|@SCALING01
kalomazeKA#836|@KALOMAZE
elieEL#716|@ELIEBAKOUCH
Bojan TunguzBT#687|@TUNGUZ
Andrew CurranAC#517|@ANDREWCURRAN_
Peter Steinberger 🦞PS#495|@STEIPETE
kacheKA#488|@YACINEMTB
Ed H. ChiEH#434|@EDCHI
Teortaxes▶️ (DeepSeek 推特🐋铁粉 2023 – ∞)T(#420|@TEORTAXESTEX
Dean W. BallDW#392|@DEANWBALL
will brownWB#339|@WILLCCBB
Jon BarronJB#261|@JON_BARRON
Logan KilpatrickLK#90|@OFFICIALLOGANK
rohan anilRA#83|@_AROHAN_
Susan ZhangSZ#58|@SUCHENZANG
Lucas Beyer (bl16)LB#55|@GIFFMANA
Nando de FreitasND#29|@NANDODF
Sasha RushSR#23|@SRUSH_NLP
Jeff DeanJD#2|@JEFFDEAN
50

A new analysis projects hundreds of billions of dollars in philanthropic capital from AI companies centered on the OpenAI Foundation’s $220 billion stake and Anthropic co-founders’ 80 percent wealth pledges — Scale matches funding for five new universities at $35 billion yearly.

1d|519.7k3.8k1.7k
Nathan is in Berkeley 🔎NI#1902|@NATHANPMYOUNG
Kim-Mai CutlerKC#1670|@KIMMAICUTLER
jessica daiJD#1653|@JESSICADAI_
Yasmin RazaviYR#1602|@YASMINRAZAVI
clare ❤️‍🔥C❤️‍🔥#1463|@CLAREJTBIRCH
Ben GolubBG#1447|@BEN_GOLUB
rohitRO#1220|@KRISHNANROHIT
José Luis Ricón Fernández de la PuenteJL#1106|@ARTIRKEL
Seth LazarSL#1060|@SETHLAZAR
Beff (e/acc)B(#839|@BEFFJEZOS
Adrien EcoffetAE#719|@ADRIENLE
Dean W. BallDW#392|@DEANWBALL
kipplyKI#371|@KIPPERRII
@timnitGebru (@dair-community.social/bsky.social)@(#195|@TIMNITGEBRU
Joshua AchiamJA#109|@JACHIAM0
60

Exa raised $250 million in a Series C at a $2.2 billion valuation led by Andreessen Horowitz, reporting 400,000 developers and 5,000 company adopters for its AI agent search platform — Token usage grew 20x for agent-driven queries.

17h|1.2M4.4k2k
andrew gaoAG#1924|@ITSANDREWGAO
Sarah WangSW#1814|@SARAHDINGWANG
mickey friedmanMF#1774|@MICKEYXFRIEDMAN
AviAV#1713|@AVISCHIFFMANN
Will BrykWB#1582|@WILLIAMBRYK
jasonJA#929|@JXNLCO
LishaLI#761|@LISHALI88
Keerthana GopalakrishnanKG#730|@KEERTHANPG
Alexis RossAR#563|@ALEXISJROSS
Marco MascorroMM#557|@MASCOBOT
Garry TanGT#270|@GARRYTAN
swyx🛬 SFOSS#214|@SWYX
Igor BabuschkinIB#117|@IBAB
Paul GrahamPG#78|@PAULG
70

SpaceX files S-1 for Nasdaq IPO under ticker SPCX targeting $1.75 trillion valuation and discloses xAI driving 60 percent of 2025 capital spending — Grok reached 550 million monthly active users with 1 GW of compute deployed.

9h|4.5M34.6k1.4k
Steven SinofskySS#1992|@STEVESI
Nick DobosND#1894|@NICKADOBOS
bilalBI#1402|@BILALTWOVEC
Lisan al GaibLA#980|@SCALING01
Erik BrynjolfssonEB#389|@ERIKBRYN
Elon MuskEM#76|@ELONMUSK
80

Investor Gavin Baker outlines wafer shortages, power availability, and DRAM supply as potential bottlenecks for AI scaling and valuations in a discussion with Patrick O'Shaughnessy — Describes orbital data centers using 3,000-pound Blackwell racks with 500-foot solar wings.

1d|2.4M6.4k5.8k
Patrick OShaughnessyPO#1862|@PATRICK_OSHAG
Matthew BermanMB#1759|@MATTHEWBERMAN
Brad GerstnerBG#1614|@ALTCAP
Shaun MaguireSM#1333|@SHAUNMMAGUIRE
Gavin BakerGB#1266|@GAVINSBAKER
Dylan PatelDP#170|@DYLAN522P
90

Jeff Bezos says artificial intelligence will elevate people at work rather than replace jobs, citing assistance in X-ray analysis and programming tasks during a CNBC interview — Bezos linked efficiency gains to potential deflationary effects and cautioned against early regulation.

17h|718.9k8.7k1.4k
Beff (e/acc)B(#839|@BEFFJEZOS
Aaron LevieAL#562|@LEVIE
Andrew CurranAC#517|@ANDREWCURRAN_
100

Anthropic projects $10.9 billion June quarter revenue and its first $559 million operating profit, implying a $44 billion annualized run rate during a funding round likely valuing it above OpenAI. — Limited compute capacity pushes some customers to other providers.

9h|377.1k2.7k439
Timothy B. LeeTB#1556|@BINARYBITS
ZephyrZE#1497|@ZEPHYR_Z9
Chubby♨️CH#1496|@KIMMONISMUS
Lisan al GaibLA#980|@SCALING01
Chris PaxtonCP#732|@CHRIS_J_PAXTON
Andrew CurranAC#517|@ANDREWCURRAN_
110

A Stanford PhD student's study finds 11 major AI models agree with users 49 percent more often than humans across nearly 12,000 social scenarios — Models endorsed lying or illegal actions 47 percent of time

21h|2.6M13.8k8.3k
Lester MackeyLM#1255|@LESTERMACKEY
Eric JangEJ#36|@ERICJANG11
120

Figma introduced an AI agent embedded directly on its design canvas that understands native interfaces and team workflows to adjust layouts, apply design systems, and refine prototypes via chat prompts — A 61-second demo analyzes the Bookworm prototype on Cloudflare infrastructure.

17h|647.9k4.3k1.9k
ben hylakBH#1941|@BENHYLAK
130

SpaceX is hiring world-class engineers and physicists for its SpaceXAI team and states that no prior artificial intelligence experience is required — Elon Musk directed applications to ai_eng@spacex.com with three bullet points.

1h|5.9M27.9k6.1k
TiboTI#768|@THSOTTIAUX
Elon MuskEM#76|@ELONMUSK
140

OpenAI announces one of its models identified new constructions for the planar unit distance problem first posed by Paul Erdős in 1946 that outperform square-grid arrangements long assumed optimal — Marks first autonomous AI solution to a longstanding open math problem.

8h|98k1.2k242
Nick DobosND#1894|@NICKADOBOS
thebesTH#1499|@VOOOOOOGEL
Samuel Hammond 🦉SH#1488|@HAMANDCHEESE
Shaun MaguireSM#1333|@SHAUNMMAGUIRE
PrakashPR#1332|@8TEAPI
Alex VolkovAV#1245|@ALTRYNE
Andrea MontanariAM#1088|@ANDREA__M
Miles BrundageMB#20|@MILES_BRUNDAGE
150

Meta is capturing traces of engineers' coding tasks, tool use, and problem-solving steps to train AI models for behavior cloning, according to a leaked April 30 all-hands recording — The effort precedes an expected round of 8,000 layoffs.

1d|875.7k2.6k1.5k
Alex ImasAI#1777|@ALEXOLEGIMAS
Chubby♨️CH#1496|@KIMMONISMUS
rohitRO#1220|@KRISHNANROHIT
Rohan PaulRP#1032|@ROHANPAUL_AI
160

Intuit is reducing its global workforce by 17 percent, or roughly 3,000 employees, to redirect resources toward AI efforts according to an internal memo reviewed by Reuters — Posts reference Meta cuts and SaaS predictions of 40-50 percent reductions.

17h|873.7k4.2k349
ZephyrZE#1497|@ZEPHYR_Z9
@jason@J#1061|@JASON
Rohan PaulRP#1032|@ROHANPAUL_AI
Andrew CurranAC#517|@ANDREWCURRAN_
170

Elon Musk, xAI founder and CEO, states the trend is strong for Composer 2.5 following a post requesting its availability in Grok Build — Beff had suggested adding the xAI system to Grok Build on X.

1d|5.8M20.8k1.7k
martin_casadoMA#470|@MARTIN_CASADO
Elon MuskEM#76|@ELONMUSK
180

Terminal-Bench Science extends the original Terminal-Bench benchmark used by Anthropic, OpenAI, and Google DeepMind into scientific domains and opens for over 100 task contributions by August 17, 2026 — Contributors package workflows as RL environments with verification tests.

13h|845.1k579224
Katie KangKK#1481|@KATIE_KANG_
Sanmi KoyejoSK#1085|@SANMIKOYEJO
Alex RatnerAR#1070|@AJRATNER
Lisan al GaibLA#980|@SCALING01
Chenhao TanCT#570|@CHENHAOTAN
Ludwig SchmidtLS#366|@LSCHMIDT3
Alex DimakisAD#172|@ALEXGDIMAKIS
Thomas WolfTW#17|@THOM_WOLF
190

DeepSeek forms a new Harness team to develop Code Harness from the ground up, opening two roles in Beijing under its 2026 social recruitment drive — Effort targets code infrastructure for internal and potential public use.

1d|364.3k1.9k488
Deli ChenDC#1523|@VICTOR207755822
ZephyrZE#1497|@ZEPHYR_Z9
Pasquale MinerviniPM#713|@PMINERVINI
Teortaxes▶️ (DeepSeek 推特🐋铁粉 2023 – ∞)T(#420|@TEORTAXESTEX
Xin Eric Wang (hiring postdoc)XE#417|@XWANG_LK
200

Pirate Wires analysis claims data centers would consume just 8 percent of U.S. golf course water usage by 2030 even after tripling their draw — 400 Arizona golf courses use 42 billion gallons yearly.

13h|665.6k4k820
Mike SolanaMS#1323|@MICSOLANA
Elon MuskEM#76|@ELONMUSK
210

Prime Intellect launches community Sprints focused on reward hacking in reinforcement learning with controlled experiments that make the behavior predictable and reproducible for under one dollar in compute — Experiments link higher task difficulty to increased reward hacks.

8h|101.3k899427
Florian BrandFB#1153|@XEOPHON
Vincent WeisserVW#707|@VINCENTWEISSER
will brownWB#339|@WILLCCBB
220

Eric Schmidt draws boos from Arizona State University graduates for every reference to artificial intelligence in his commencement address — Similar reactions hit other recent U.S. graduation speeches amid job loss fears.

2d|250.9k791290
Daniel JeffriesDJ#1988|@DAN_JEFFRIES1
@jason@J#1061|@JASON
Rohan PaulRP#1032|@ROHANPAUL_AI
Gary MarcusGM#153|@GARYMARCUS
230

Tuhin Chakrabarty and Sam Rodriques reject claims that three recent Nature papers show AI systems fully replicating core scientific tasks and rendering scientists obsolete — Rodriques says human judgment remains essential for five-to-ten-year productivity gains.

1d|142.2k1k413
Anshul KundajeAK#1675|@ANSHULKUNDAJE
Anders SandbergAS#1054|@ANDERSSANDBERG
Sam RodriquesSR#1007|@SGRODRIQUES
Andrew CurranAC#517|@ANDREWCURRAN_
Nathan BenaichNB#242|@NATHANBENAICH
240

Junyeob Baek and colleagues introduce Generative Recursive reAsoning Models (GRAM) that convert deterministic recursive reasoning into stochastic latent trajectories and report 97.0% accuracy on Sudoku-Extreme with 10 million parameters — The model also scores 52.0% on ARC-AGI-1 and 44.6% on ARC-AGI-2.

17h|158.1k1.4k1.2k
Super DarioSD#1776|@INDUCTIONHEADS
Kyle KastnerKK#1025|@KASTNERKYLE
Mengye RenMR#880|@MENGYER
billy bubbaBB#840|@WGRATHWOHL
250

The Economist attributes most recent US GDP growth to AI-hardware investment, with that category indexed above 140 by 2026 while overall GDP stays near 100. — Separate note shows AI fixed investment at 5.46 percent of GDP.

15h|113.1k1.3k421
Nathan is in Berkeley 🔎NI#1902|@NATHANPMYOUNG
Alex ImasAI#1777|@ALEXOLEGIMAS
Adrien EcoffetAE#719|@ADRIENLE
260

Sam Altman, co-founder and CEO of OpenAI, identifies three AGI priority areas of accelerating research, companies, and personal goal achievement while referencing a unit distance result and $2 million credits for YC companies — Reply asks if personal AGI could build a moon base.

8h|364k5.9k683
Beff (e/acc)B(#839|@BEFFJEZOS
Jimmy Apples 🍎/accJA#837|@APPLES_JIMMY
Sam AltmanSA#8|@SAMA
270

An AI system solves a widely known open mathematics problem that multiple leading human mathematicians had been unable to resolve after significant effort — Mathematician Timothy Gowers called it one of AI's firsts in the field.

10h|42.4k54874
Timothy B. LeeTB#1556|@BINARYBITS
Shubhendu TrivediST#1446|@_ONIONESQUE
bilalBI#1402|@BILALTWOVEC
PrakashPR#1332|@8TEAPI
Kevin Weil 🇺🇸KW#451|@KEVINWEIL
Csaba SzepesvariCS#409|@CSABASZEPESVARI
Michael BronsteinMB#199|@MMBRONSTEIN
Thang LuongTL#188|@LMTHANG
280

The White House plans to release an AI executive order as early as tomorrow directing agencies to create a voluntary clearinghouse for AI vulnerabilities and a classified benchmark process for frontier models — Tuesday briefing covered the provisions with OpenAI, Anthropic and Reflection AI.

1d|180.6k691155
Bindu ReddyBR#1622|@BINDUREDDY
Stephanie PalazzoloSP#1166|@STEPH_PALAZZOLO
Beff (e/acc)B(#839|@BEFFJEZOS
Andrew CurranAC#517|@ANDREWCURRAN_
Miles BrundageMB#20|@MILES_BRUNDAGE
290

SimWorld releases SimWorld Studio with SimCoder, generating interactive 3D environments on Unreal Engine 5 that raise embodied navigation success rates from 50% to 90% — Includes GitHub repository, arXiv paper, and urban demo assets.

12h|96.9k305193
Zhiting HuZH#949|@ZHITINGHU
Lianhui QinLQ#737|@LIANHUIQ
300

METR evaluations find frontier AI agents rely on explicit natural language chain-of-thought to complete hardest tasks, with time horizons dropping from 1.5–2 years to about 4 minutes when actions must stay hidden — David Rein from METR spent a month stress-testing controls at Anthropic.

1d|41.4k330129
Samuel Hammond 🦉SH#1488|@HAMANDCHEESE
gavin leech (Non-Reasoning)GL#1480|@GLEECH
Toby OrdTO#1159|@TOBYORDOXFORD
Maksym AndriushchenkoMA#1066|@MAKSYM_ANDR
Peter HasePH#913|@PETERBHASE
⿻ Andrew Trask⿻A#361|@IAMTRASK
Gary MarcusGM#153|@GARYMARCUS