Digg - AI news, before it trends

Today's Highlights

In case you missed it

OpenAI reports an internal reasoning model identified new super-linear constructions for Erdős's planar unit distance problem from 1946

CS#43|@CHRSZEGEDY

EJ#36|@ERICJANG11

NB#31|@POLYNOAMIAL

MB#20|@MILES_BRUNDAGE

#1 VIEWED

Cursor, the AI code editor from Anysphere, doubles usage limits for the next month for new users invited onto its Teams plan

RL#1733|@RYOLU_

MT#506|@MNTRUELL

EM#76|@ELONMUSK

#1 COMMENTS

Sam Altman, OpenAI co-founder and CEO, asks what problem people most hope AI will solve in the future

TA#1047|@VICTORTAELIN

B(#839|@BEFFJEZOS

JA#837|@APPLES_JIMMY

KA#488|@YACINEMTB

TM#359|@ISCIENCELUVR

#1 BOOKMARKED

8.7k

Microsoft ends internal use of Anthropic's Claude coding tool after token-based billing costs prove unsustainable despite its resources

YB#522|@YAROSLAVVB

GM#153|@GARYMARCUS

TW#17|@THOM_WOLF

FASTEST CLIMBING

Google's Gemini Omni model generates first-person video of a taxi following a route from a marked Google Maps screenshot with consistent viewpoint and visual details

NR#1562|@NATANIELRUIZG

BI#1402|@BILALTWOVEC

JB#261|@JON_BARRON

RISING LIKES

944

OpenAI implements an internal rule banning vagueposting on social media and adds a process for staff to report violations by colleagues

𝚟⟢#1753|@VIEMCCOY

JJ#244|@JOANNEJANG

Top Stories

Posts:—|Clusters:—|Next Crawl:—

10

Alibaba's Qwen team releases Qwen3.7-Max, its closed-weights flagship model for agent workflows that completes long-horizon tasks like 35-hour kernel optimizations with over 1,000 tool calls — Scored 56.6 on Artificial Analysis Intelligence Index after 4.8-point gain.

16h|877.4k7.3k1.3k

SD#1776|@INDUCTIONHEADS

ZE#1497|@ZEPHYR_Z9

CH#1496|@KIMMONISMUS

FB#1153|@XEOPHON

RP#1032|@ROHANPAUL_AI

LA#980|@SCALING01

AC#517|@ANDREWCURRAN_

T(#420|@TEORTAXESTEX

T🪽#259|@TEKNIUM

JH#40|@JEREMYPHOWARD

21

Sam Altman, OpenAI co-founder and CEO, asks what problem people most hope AI will solve in the future — Replies suggest curing diseases and automating research.

6h|1.4M9.2k1k#1COMMENTS10.1k

MB#1759|@MATTHEWBERMAN

CH#1496|@KIMMONISMUS

NC#1319|@CHOWDHURYNEIL

SS#1073|@SYTELUS

TA#1047|@VICTORTAELIN

B(#839|@BEFFJEZOS

JA#837|@APPLES_JIMMY

KA#488|@YACINEMTB

TM#359|@ISCIENCELUVR

31

Microsoft ends internal use of Anthropic's Claude coding tool after token-based billing costs prove unsustainable despite its resources — Uber CTO reports exhausting full 2026 AI budget in four months.

10h|4.6M15.7k#1BOOKMARKED8.7k

DJ#1988|@DAN_JEFFRIES1

AI#1777|@ALEXOLEGIMAS

YB#522|@YAROSLAVVB

GM#153|@GARYMARCUS

TW#17|@THOM_WOLF

40

Stanford researchers release paper 'A Bitter Lesson for Data Filtering' showing large models improve without data filtering in high-compute data-scarce regimes — Work targets abundant compute and limited data pretraining scenarios.

14h|145.4k1.2k758

AK#1675|@ANSHULKUNDAJE

JW#1460|@JIAXINWEN22

Z"#1381|@WZENUS

AP#1365|@PANDAASHWINEE

SS#1073|@SYTELUS

CL#442|@CHENGLEISI

XM#437|@INFOXIAO

XE#417|@XWANG_LK

TH#231|@TATSU_HASHIMOTO

MB#199|@MMBRONSTEIN

SX#158|@SAININGXIE

LB#55|@GIFFMANA

50

Turbopuffer reached a $100 million annual run rate in March, 19 months after hitting the $1 million mark, while remaining profitable on less than $1 million in total funding — Its object storage database first powered Cursor and now serves Anthropic and Notion.

15h|1.4M6.1k1.7k

BH#1941|@BENHYLAK

SW#1814|@SARAHDINGWANG

LG#1030|@LACHYGROOM

CB#999|@CODE_STAR

EB#746|@BERNHARDSSON

KA#488|@YACINEMTB

New github stars (48hrs)

st3v3nmw/obsidian-spaced-repetition2.4k8m

Schedules spaced-repetition reviews for flashcards and notes inside Obsidian.

aisa-group/InferenceBench131d

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

60

ClickUp announces 22% workforce reduction to restructure around a 100x AI-driven organization with AI-centric workflows — CEO DJ Curfew says the business has never been stronger.

11h|962.9k2.3k1.2k

ND#1894|@NICKADOBOS

AF#1429|@SOLOGEN

JH#1281|@MATHEMAGIC1AN

LR#1176|@LENNYSAN

RS#612|@ZIV_RAVID

JH#40|@JEREMYPHOWARD

70

Aleksander Madry, a machine learning researcher at OpenAI and MIT professor on leave, announced his departure after nearly three years heading preparedness and safety efforts — He was reassigned last summer to a role focused on AI reasoning.

11h|129.1k1.2k219

NR#1562|@NATANIELRUIZG

SM#487|@SANGMICHAELXIE

KW#451|@KEVINWEIL

AK#354|@ANANYAKU

GK#210|@THEGAUTAMKAMATH

BB#133|@BOAZBARAKTCS

AM#93|@ALEKS_MADRY

MC#71|@MARKCHEN90

DR#56|@ROYDANROY

80

OpenAI adds Appshots feature to Codex that lets Mac users press Command-Command to capture and attach screenshots of open windows to active threads — Demo shows integration with Google Docs, recipes, and calendars.

11h|526.7k4k679

CH#1496|@KIMMONISMUS

SW#801|@SHERWINWU

LK#627|@LUKASZKAISER

90

Cursor, the AI code editor from Anysphere, doubles usage limits for the next month for new users invited onto its Teams plan — CEO Michael Truell announced the offer on X alongside new feature references.

14h|#1VIEWS96.5M#1LIKED23.5k1.2k

RL#1733|@RYOLU_

MT#506|@MNTRUELL

EM#76|@ELONMUSK

100

AI Security Institute published a report on May 21, 2026, assessing oversight mechanisms for advanced AI systems against rapid capability gains and identifying degradation pathways. — Report section records disputed advantages of discrete token reasoning.

13h|38.3k282127

JW#1460|@JIAXINWEN22

1A#1380|@1A3ORN

TK#1178|@TOMEKKORBAK

JW#979|@J_ASMINEWANG

C(#587|@STEPHENLCASPER

GI#347|@GEOFFREYIRVING

MB#20|@MILES_BRUNDAGE

110

ChatGPT adds a direct integration with Microsoft PowerPoint through a new app that lets users generate slides, query deck content, and apply edits without leaving the application — Built by developer Ryan Brewer at chatgpt.com/apps/powerpoint.

10h|406.1k3.7k960

NT#1004|@NICKATURLEY

AK#970|@AMINKARBASI

BP#357|@BORISMPOWER

122

Greg Brockman, OpenAI president and co-founder, says the model alone is no longer the product in a post addressing standalone AI offerings — Victor Taelin replied that the claim is wrong and signaled he is shorting.

2h|127k2.1k168

HH#1277|@HAMELHUSAIN

TA#1047|@VICTORTAELIN

HB#1022|@HERBIEBRADLEY

EM#175|@EMOLLICK

130

GoodfireAI shares research showing sparse autoencoders tile and shatter curved neural manifolds in large models rather than linear directions, recasting unsupervised discovery as an inverse Ising problem — Visualizations map features across a manifold spanning 1800 to 1998.

14h|133.8k1.4k784

TM#1626|@BANBURISMUS_

CF#1356|@CFGEEK

MG#1329|@MEGAMOR2

KD#908|@CSPROFKGD

142

Hark raises over $700 million in Series A funding at a $6 billion post-money valuation to scale its AI foundation models and custom hardware — Round led by Parkway Venture Capital with NVIDIA and AMD participating

16h|492.7k2.4k609

B(#839|@BEFFJEZOS

BA#688|@ADCOCK_BRETT

RS#321|@SCOBLEIZER

150

Every publishes 'After Automation' report detailing Codex and Claude Code automation of coding, writing, design and customer service as staff grows from 4 to 30 — Every co-founder Dan Shipper notes AI raised demand for human work.

12h|771.2k1.6k2.4k

DJ#1988|@DAN_JEFFRIES1

MB#1759|@MATTHEWBERMAN

DS#1410|@DANSHIPPER

LR#1176|@LENNYSAN

160

OpenAI enabled a locked use feature for Codex that permits the system to control Mac applications from a connected phone even when the device is locked and the screen is off — It requires an Apple authorization plug-in for macOS sessions.

11h|1M7.4k2k

TI#768|@THSOTTIAUX

170

Gated DeltaNet-2 separates channel-wise erase and write gates within linear attention, raising S-NIAH-3 scores from 63 to 90 on 1.3B models trained on 100B tokens — The mechanism appears in Qwen3-Next and Qwen3.5 397B models.

7h|118.1k834473

EL#716|@ELIEBAKOUCH

T(#420|@TEORTAXESTEX

YC#52|@YEJINCHOINKA

180

Roon from OpenAI says the upbeat PBS Kids-style tone of new AI release videos mismatches the technology and carries Faustian implications unlike Carl Sagan's explanatory style — Samo Burja quoted the post and Victor Taelin urged matching ads.

9h|61.6k1k116

TA#1047|@VICTORTAELIN

VW#707|@VINCENTWEISSER

WD#254|@WILLDEPUE

191

Modal scales its serverless AI infrastructure to manage rising demand while extending support for model post-training workloads — Eric Zhang, former Modal founding engineer, highlighted the updates.

9h|85.3k65853

HH#1277|@HAMELHUSAIN

EZ#1068|@EKZHANG1

C🎉#863|@CHARLES_IRL

EB#746|@BERNHARDSSON

SC#733|@SARAHCAT21

201

OpenAI valuation rose 67 percent and Anthropic valuation rose 173 percent since the beginning of the year in posts referencing a February 23 Marcus on AI Substack article. — Anthropic reported second-quarter profits alongside OpenAI's Erdős problem solution.

18h|133.1k1.3k170

SD#1776|@INDUCTIONHEADS

LA#980|@SCALING01

GM#153|@GARYMARCUS

Rising Stories

Grok Update5H AGO
Grok Build Adds X Search Integration and Faster Web Search
#341@BILLYUCHENLIN
#76@ELONMUSK
11k15.3M
Policy Move3H AGO
David Sacks Calls Trump to Postpone AI Executive Order
#1759@MATTHEWBERMAN
#1616@NITASHATIKU
14449.2k
AI Research2H AGO
Equilibrium Reasoners reach near-perfect accuracy on extreme reasoning benchmarks
#538@RUIQIGAO
#67@ZICOKOLTER
23334.9k
AI Debate1H AGO
Gary Marcus moves up his AI bubble collapse prediction to days or weeks
#931@VENTURETWINS
#59@TSZZL
70822.4k
Internal Rule2H AGO
OpenAI bans vagueposting on social media for employees
#244@JOANNEJANG
#59@TSZZL
91631.9k

Github Stars

(7 days)

…ilsonnn/image-blaster5D AGO AGO
Converts an input image into 3D meshes, Gaussian splats, and audio files using Claude skills and external AI APIs.
1673
475
878
3.1k stars
…ian-spaced-repetition8M AGO AGO
Schedules spaced-repetition reviews for flashcards and notes inside Obsidian.
681
484
2.4k stars
…-group/InferenceBench1D AGO AGO
Evaluates autonomous AI agents optimizing LLM serving like vLLM in open-ended scenarios with quality and integrity gates.
1224
1066
13 stars
manaflow-ai/cmux2D AGO AGO
Integrates Ghostty into a macOS terminal with vertical tabs, sidebar metadata, OSC notifications, and in-app browser for AI coding agents.
203
339
17.6k stars
sapientinc/HRM-Text2D AGO AGO
Pretrains 1B HRM text models with hierarchical reasoning, task completion, and low-compute PrefixLM training.
1047
707
565 stars

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

Yesterday's Top Stories, May 20, 2026.

Posts: 4,298|Clusters: 1,081|Stories frozen at midnight PT

10

OpenAI reports an internal reasoning model identified new super-linear constructions for Erdős's planar unit distance problem from 1946 — The model used algebraic number theory methods without math-specific training.

11h|#1VIEWS12.6M#1LIKED63k#1BOOKMARKED14.7k

SW#801|@SHERWINWU

EA#757|@AKYUREKEKIN

BM#741|@MCKBRANDO

CP#732|@CHRIS_J_PAXTON

EL#716|@ELIEBAKOUCH

TB#699|@TAMAYBES

JL#677|@JERRYJLIU0

JP#672|@ZHANSHENG

PD#654|@PMDDOMINGOS

YS#638|@YONASHAV

SR#604|@CLONEOFSIMO

HE#572|@HEINRICHKUTTLER

CC#548|@ITSCLIVETIME

AC#517|@ANDREWCURRAN_

CL#442|@CHENGLEISI

T(#420|@TEORTAXESTEX

DW#392|@DEANWBALL

AW#367|@ALEXWEI_

BP#357|@BORISMPOWER

NM#345|@__NMCA__

FF#331|@FRANCOISFLEURET

AC#263|@ANDREW_N_CARR

SG#256|@SURYAGANGULI

WD#254|@WILLDEPUE

AC#241|@_AIDAN_CLARK_

JP#206|@MERETTM

MB#199|@MMBRONSTEIN

MN#187|@MICHAEL_NIELSEN

EM#175|@EMOLLICK

AD#172|@ALEXGDIMAKIS

EM#167|@EMOSTAQUE

AG#154|@ANDREWGWILS

GM#153|@GARYMARCUS

BB#133|@BOAZBARAKTCS

DR#103|@DELIPRAO

OP#72|@OFIRPRESS

DR#56|@ROYDANROY

LB#55|@GIFFMANA

SB#51|@SEBASTIENBUBECK

CS#43|@CHRSZEGEDY

EJ#36|@ERICJANG11

NB#31|@POLYNOAMIAL

MB#20|@MILES_BRUNDAGE

20

SpaceX's IPO filing discloses a $1.25 billion monthly compute deal with Anthropic for Colossus clusters through May 2029 — Capacity ramps in May 2026 for AI training including Grok 5.

14d|11.6M61.9k3.9k

GL#1984|@GARRISONLOVELY

PO#1862|@PATRICK_OSHAG

🍓🍓#1711|@IRULETHEWORLDMO

NT#1616|@NITASHATIKU

AA#1540|@AUSTEN

ZE#1497|@ZEPHYR_Z9

CH#1496|@KIMMONISMUS

BI#1402|@BILALTWOVEC

SM#1333|@SHAUNMMAGUIRE

RM#1327|@RACHELMETZ

SI#1124|@SIGNULLL

KR#1046|@KYLIEBYTES

LA#980|@SCALING01

AD#867|@DORIALEXANDER

B(#839|@BEFFJEZOS

EL#716|@ELIEBAKOUCH

ZN#537|@ZACHARYNADO

AC#517|@ANDREWCURRAN_

T(#420|@TEORTAXESTEX

TB#384|@NOTTOMBROWN

TM#359|@ISCIENCELUVR

RS#321|@SCOBLEIZER

EM#76|@ELONMUSK

LB#55|@GIFFMANA

30

Google's Gemini 3.5 Flash model tops Zapier's Automation Bench at 14.5 percent and leads the APEX-Agents leaderboard at 47.1 percent — It runs 4x faster than Gemini 3.1 Pro on coding tasks.

2d|842.1k9.1k755

ET#1884|@TORCHCOMPILED

BR#1622|@BINDUREDDY

AV#1245|@ALTRYNE

SI#1124|@SIGNULLL

RP#1032|@ROHANPAUL_AI

CB#999|@CODE_STAR

B(#982|@ANDERSONBCDEFG

LA#980|@SCALING01

KA#836|@KALOMAZE

EL#716|@ELIEBAKOUCH

AC#517|@ANDREWCURRAN_

PS#495|@STEIPETE

KA#488|@YACINEMTB

T(#420|@TEORTAXESTEX

DW#392|@DEANWBALL

WB#339|@WILLCCBB

JB#261|@JON_BARRON

LK#90|@OFFICIALLOGANK

RA#83|@_AROHAN_

SZ#58|@SUCHENZANG

LB#55|@GIFFMANA

SR#23|@SRUSH_NLP

40

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

ET#1884|@TORCHCOMPILED

ST#1629|@STOCHASTICCHASM

WH#1430|@NREHIEW_

NR#1167|@NILS_REIMERS

FB#1153|@XEOPHON

B(#982|@ANDERSONBCDEFG

EL#716|@ELIEBAKOUCH

JA#715|@JAYALAMMAR

PM#713|@PMINERVINI

ME#675|@MERVENOYANN

NF#601|@NICKFROSST

KA#488|@YACINEMTB

EL#475|@OMARSAR0

T(#420|@TEORTAXESTEX

SR#370|@SIVAREDDYG

WB#339|@WILLCCBB

MG#258|@MARCGBELLEMARE

JP#216|@JPINEAU1

SB#208|@BLANCHEMINERVA

AG#171|@AIDANGOMEZ

DR#103|@DELIPRAO

C🤗#68|@CLEMENTDELANGUE

AK#28|@_AKHALIQ

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

NI#1902|@NATHANPMYOUNG

KC#1670|@KIMMAICUTLER

JD#1653|@JESSICADAI_

YR#1602|@YASMINRAZAVI

C❤️‍🔥#1463|@CLAREJTBIRCH

BG#1447|@BEN_GOLUB

RO#1220|@KRISHNANROHIT

JL#1106|@ARTIRKEL

SL#1060|@SETHLAZAR

B(#839|@BEFFJEZOS

AE#719|@ADRIENLE

DW#392|@DEANWBALL

KI#371|@KIPPERRII

@(#195|@TIMNITGEBRU

JA#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

AG#1924|@ITSANDREWGAO

SW#1814|@SARAHDINGWANG

MF#1774|@MICKEYXFRIEDMAN

AV#1713|@AVISCHIFFMANN

WB#1582|@WILLIAMBRYK

LI#761|@LISHALI88

KG#730|@KEERTHANPG

AR#563|@ALEXISJROSS

MM#557|@MASCOBOT

GT#270|@GARRYTAN

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

SS#1992|@STEVESI

ND#1894|@NICKADOBOS

BI#1402|@BILALTWOVEC

LA#980|@SCALING01

EB#389|@ERIKBRYN

EM#76|@ELONMUSK

80

Investor Gavin Baker examines wafer shortages and power limits as constraints on AI scaling along with orbital data centers and Elon Musk's Terafab project — Episode also covers Cerebras IPO and TSMC foundry strategies.

1d|2.4M6.4k5.8k

PO#1862|@PATRICK_OSHAG

MB#1759|@MATTHEWBERMAN

BG#1614|@ALTCAP

SM#1333|@SHAUNMMAGUIRE

GB#1266|@GAVINSBAKER

DP#170|@DYLAN522P

90

OpenAI solved the unit distance problem in mathematics, prompting discussion on whether the work used informal chain-of-thought reasoning or neurosymbolic tools such as Lean — Attention now turns to AI progress in social science domains.

3d|252.9k1.6k169

TB#1556|@BINARYBITS

ST#1446|@_ONIONESQUE

BI#1402|@BILALTWOVEC

PR#1332|@8TEAPI

LA#980|@SCALING01

AD#867|@DORIALEXANDER

RS#612|@ZIV_RAVID

KW#451|@KEVINWEIL

CS#409|@CSABASZEPESVARI

AK#228|@BLACKHC

MB#199|@MMBRONSTEIN

TL#188|@LMTHANG

GM#153|@GARYMARCUS

100

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

B(#839|@BEFFJEZOS

AC#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

22h|2.6M13.8k8.3k

LM#1255|@LESTERMACKEY

EJ#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

BH#1941|@BENHYLAK

130

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.

10h|377.1k2.7k439

TB#1556|@BINARYBITS

ZE#1497|@ZEPHYR_Z9

CH#1496|@KIMMONISMUS

LA#980|@SCALING01

CP#732|@CHRIS_J_PAXTON

AC#517|@ANDREWCURRAN_

140

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

TI#768|@THSOTTIAUX

EM#76|@ELONMUSK

150

Meta captures traces of its engineers' internal coding tasks, tool usage, and keystrokes to train AI models via behavior cloning, according to a leaked April 30 all-hands meeting — The effort precedes planned cuts to roughly 8,000 roles.

1d|875.7k2.6k1.5k

AI#1777|@ALEXOLEGIMAS

CH#1496|@KIMMONISMUS

RO#1220|@KRISHNANROHIT

RP#1032|@ROHANPAUL_AI

160

OpenAI will provide $2 million in tokens to every startup in the current Y Combinator batch in exchange for equity — Sam Altman presented the offer at a Y Combinator event.

1d|379k6.4k690

B(#839|@BEFFJEZOS

JA#837|@APPLES_JIMMY

170

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

ZE#1497|@ZEPHYR_Z9

RP#1032|@ROHANPAUL_AI

AC#517|@ANDREWCURRAN_

180

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

MA#470|@MARTIN_CASADO

EM#76|@ELONMUSK

190

White House postpones signing ceremony for executive order on AI and cybersecurity risks from advanced models after review of provisions that could slow US companies — Axios cited internal administration disagreements over the order.

1d|187.9k735165

BR#1622|@BINDUREDDY

SH#1488|@HAMANDCHEESE

SP#1166|@STEPH_PALAZZOLO

B(#839|@BEFFJEZOS

AC#517|@ANDREWCURRAN_

MB#20|@MILES_BRUNDAGE

200

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

KK#1481|@KATIE_KANG_

SK#1085|@SANMIKOYEJO

AR#1070|@AJRATNER

LA#980|@SCALING01

CT#570|@CHENHAOTAN

LS#366|@LSCHMIDT3

AD#172|@ALEXGDIMAKIS

TW#17|@THOM_WOLF

210

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

DC#1523|@VICTOR207755822

ZE#1497|@ZEPHYR_Z9

PM#713|@PMINERVINI

T(#420|@TEORTAXESTEX

XE#417|@XWANG_LK

220

Pangram Labs analysis flags AI-generated text in three 2026 Commonwealth Short Story Prize regional winners plus the 2025 overall winner and a Granta-published story — Literary foundation and Granta maintain the entries are human-written.

2d|81.8k810148

GL#1480|@GLEECH

RO#1220|@KRISHNANROHIT

TC#1050|@TUHINCHAKR

🎭#878|@DEEPFATES

X(#745|@XUANALOGUE

ZN#537|@ZACHARYNADO

NS#227|@NSAPHRA

230

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.

14h|665.6k4k820

MS#1323|@MICSOLANA

EM#76|@ELONMUSK

240

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

FB#1153|@XEOPHON

VW#707|@VINCENTWEISSER

WB#339|@WILLCCBB

250

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

DJ#1988|@DAN_JEFFRIES1

RP#1032|@ROHANPAUL_AI

GM#153|@GARYMARCUS

260

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

AK#1675|@ANSHULKUNDAJE

AS#1054|@ANDERSSANDBERG

SR#1007|@SGRODRIQUES

AC#517|@ANDREWCURRAN_

NB#242|@NATHANBENAICH

270

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

SD#1776|@INDUCTIONHEADS

KK#1025|@KASTNERKYLE

MR#880|@MENGYER

BB#840|@WGRATHWOHL

280

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

NI#1902|@NATHANPMYOUNG

AI#1777|@ALEXOLEGIMAS

AE#719|@ADRIENLE

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

ZH#949|@ZHITINGHU

LQ#737|@LIANHUIQ

300

Antigravity unifies its agentic surfaces under a single platform incorporating Antigravity 2.0 desktop app, CLI, SDK, and IDE for consistent access across environments — Demo builds Street Guesser game using multi-agent Gemini interfaces.

1d|534.9k2.7k429

BH#1941|@BENHYLAK

PS#495|@STEIPETE