Tech News

Today's Highlights

In case you missed it

Anysphere launches Cursor for iOS, letting developers run always-on cloud agents and manage desktop agents remotely

TS#1050|@TIMSWEENEYEPIC

MM#716|@MASCOBOT

B(#530|@BEFFJEZOS

MA#315|@MARTIN_CASADO

EM#52|@ELONMUSK

It’s a new day at Digg. Fresh stories are clustering now. 30% complete.

Top Stories

Posts:—|Clusters:—|Next Crawl:—

10

DeepSeek releases DSpark, a speculative decoding algorithm boosting DeepSeek-V4 throughput by up to 661%Open-source vLLM has rapidly integrated support for the algorithm.

3h|4.1k113

ZE#1695|@ZEPHYR_Z9

T(#501|@TEORTAXESTEX

KA#403|@YACINEMTB

Top post: @zephyr_z9 “DSpark will see crazy adoption The Whale reduced the compute pressure on Chinese AI infra”

20

Alibaba's Qwen team disputes researcher Tianhang Zhu's claims of leading its reinforcement learning developmentZhu's Google Scholar profile listed papers lacking his name.

5h|3.9k494

T(#501|@TEORTAXESTEX

SZ#84|@SUCHENZANG

Top post: @teortaxesTex “The plot thickens! I guess Tianhang can boast of his RL contributions very legitimately. PPO back then was a really high-risk, high-effort choice. I apologize.”

30

DeepSeek open-sources core ML infrastructure projects DeepEP and DeepGEMM to optimize throughput and lower inference costsNvidia's Dynamo infrastructure reportedly responds directly to these releases

1h|2.7k72

ZE#1695|@ZEPHYR_Z9

T(#501|@TEORTAXESTEX

Top post: @zephyr_z9 “Nvidia Dynamo exists because of the whale”

Watchlist

Video Signals

MN#1300|@MATTNIESSNER

FaceAnything Releases Code For High-Fidelity 4D Face Reconstruction From Videos

47m ago|Views 281Likes 11Bookmarks 10

LY#1858|@LINDAYAX

Yaccarino Links AI Adoption With GLP-1 Drugs Driving Healthcare Deflation

1h ago|Views 5.2KLikes 18Bookmarks 2

AP#1222|@APOMPLIANO

Investor Avi Felman Sells All Bitcoin to Pursue AI and Defense Tech

34m ago|Views 7.2KLikes 4Bookmarks 2

FO#1853|@FOFRAI

JP#774|@JPARKERHOLDER

Google DeepMind Opens New London Office Platform 37 Honoring AlphaGo

3h ago|Views 79KLikes 14Bookmarks 1

RP#1257|@ROHANPAUL_AI

Aravind Srinivas Warns Export Controls Accelerate China's AI Infrastructure Edge

5h ago|Views 12KLikes 3Bookmarks 0

RV|@RONALD_VANLOON

Headcam Footage from Indian Workers Trains AI to Automate Factory and Domestic Jobs

2h ago|Views 657Likes 3Bookmarks 0

AD#729|@AJDDAVISON

ReorientBot Achieves Precise Object Rearrangement and Reverse Rearrangement

43m ago|Views 75Likes 2Bookmarks 0

RV|@RONALD_VANLOON

AI Moves To Edge To Eliminate Latency In Real-Time Environments

2h ago|Views 1.1KLikes 1Bookmarks 0

RV|@RONALD_VANLOON

JCBOT Wheeled Humanoid Robot Blends Speed AI And Home-Friendly Design

22m ago|Views 316Likes 0Bookmarks 0

4

AI developer teortaxesTex argues China must spend $1.5 trillion annually to match US AI resource intensityHardware costs cap China's data center savings at 10%.

23m|2.1k72

ZE#1695|@ZEPHYR_Z9

T(#501|@TEORTAXESTEX

Top post: @zephyr_z9 “Memory+Compute+Networking is 80% of total CAPEX (China has no advantage here) The rest 20% is power+cooling+shell Assuming China is 50% cheaper compared to US The max net savings is 10%”

51

Google DeepMind opens its new London headquarters, named Platform 37 after AlphaGo's historic Move 37DeepMind staff shared first-day tours of the terraced workspace.

3h|913171

FO#1853|@FOFRAI

JP#774|@JPARKERHOLDER

Top post: @jparkerholder “Even more inspiring, it’s coming home (personal view only 🤣)”

61

ETH Zürich's Florian Tramèr questions ACL 2026 conference budget allocating $50 per coffee break with no actual mealsOrganizers claim food costs will exceed $800 per attendee.

1h|14710

FT#1659|@FLORIAN_TRAMER

GK#197|@THEGAUTAMKAMATH

Top comment: @zouharvi “@florian_tramer @aclmeeting Maybe it's because of all the vegans pumping up the catering costs? For example at last EACL we were provided with the ultimate catering experience: a plastic bowl of stale fruits (for coffee break snacks and lunch).”

Recent Stars

lijayuTnT/ShutterMuse7H AGO
Fine-tunes an MLLM to output photographer composition decisions and subject pose keypoints for capture-time guidance.
484
49 stars
mikefarah/yq9H AGO
Processes YAML, JSON, XML, CSV and other structured files using jq-like expressions on the command line.
90
15.6k stars
…mbamramfan/ai-compass9H AGO
Presents an interactive quiz scoring user views on AI along two political axes.
201
2 stars
google-deepmind/simply10H AGO
Provides minimal scalable JAX codebase for rapid iteration on LLM and autoregressive model research with AI agent support.
1194
536 stars
ACEsuit/mace11H AGO
Implements higher-order equivariant message passing neural networks for machine learning interatomic potentials.
1200
1.3k stars

Github Stars

(7 days)

xlang-ai/OSWorld-V21D AGO
Benchmarks computer-use agents on long-horizon real-world desktop tasks via VM environments and gated task sets.
1014
397
843
96 stars
deepseek-ai/DeepSpec2D AGO
Provides full-stack code for training/evaluating speculative decoding draft models with data prep and eval scripts.
484
1194
1136
2.5k stars
THUDM/slime3D AGO
Integrates Megatron training with SGLang rollouts for LLM post-training and RL scaling.
458
1014
484
7k stars
…nd-models-of-ai-tools4D AGO
Aggregates leaked and open-sourced system prompts plus tool definitions from AI coding agents like Cursor, Claude, and Devin.
1014
827
141.3k stars
anthropics/skills4D AGO
Hosts collections of self-contained skills, specs, and templates for dynamically extending Claude's capabilities on specialized tasks.
1014
1467
155.8k stars

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

Yesterday's Top Stories, Jun 29, 2026.

Posts: 3,758|Clusters: 1,274|Stories frozen at midnight PT

10

Anysphere launches Cursor for iOS, letting developers run always-on cloud agents and manage desktop agents remotelyThe mobile app supports voice dictation for background code generation

13h|2.9M14.1k#1BOOKMARKED2.2k

EZ#1911|@ERICZAKARIASSON

BR#1861|@BRIANROEMMELE

RL#1733|@RYOLU_

HH#1682|@HAMELHUSAIN

ND#1653|@NICKADOBOS

CH#1360|@KIMMONISMUS

SO#1349|@SOLEIO

SW#1092|@SJWHITMORE

TS#1050|@TIMSWEENEYEPIC

MM#716|@MASCOBOT

B(#530|@BEFFJEZOS

MA#315|@MARTIN_CASADO

EM#52|@ELONMUSK

JM

Digg Deeper question: “Is this a paid app?”

answerview answer

20

Meta AI open-sources Brain2Qwerty v2, a non-invasive brain-to-text decoder achieving 61% word accuracyIt transitions from character decoding to full-sentence semantic decoding.

17h|333.2k2.9k702

BR#1861|@BRIANROEMMELE

XL#1855|@XLR8HARDER

CB#1428|@CHRISJBAKKE

CH#1360|@KIMMONISMUS

CS#1057|@CSHORTEN30

AM#1027|@ADAMMARBLESTONE

ZN#983|@ZACHARYNADO

DF#756|@DANIELLEFONG

EL#684|@OMARSAR0

TM#613|@ISCIENCELUVR

B(#530|@BEFFJEZOS

AM#437|@ANJNEYMIDHA

AW#74|@ALEXANDR_WANG

30

Meituan releases LongCat-2.0, a 1.6-trillion-parameter MoE model with a 1-million-token context window for agentic codingThe model was trained on 50,000 Huawei Ascend 910C processors

12h|46.5k492143

LH#1586|@OHLENNART

EL#1136|@ELIEBAKOUCH

ME#861|@MERVENOYANN

TM#613|@ISCIENCELUVR

YJ#586|@YUCHENJ_UW

WB#573|@WILLCCBB

T(#501|@TEORTAXESTEX

AC#484|@ANDREW_N_CARR

KA#403|@YACINEMTB

EM#171|@EMOSTAQUE

JH#90|@JEREMYPHOWARD

AK#33|@_AKHALIQ

40

Anthropic CEO Dario Amodei argues open-source AI cannot be monitored, revoked, or patched for safety after releaseThe clip reignited debate over open-source AI safety risks

23h|176.8k1.5k142

BR#1861|@BRIANROEMMELE

CH#1360|@KIMMONISMUS

RP#1257|@ROHANPAUL_AI

ME#861|@MERVENOYANN

B(#530|@BEFFJEZOS

KA#403|@YACINEMTB

RA#218|@AGARWL_

50

Secretary of War Pete Hegseth appoints Marc Andreessen and Blake Masters to new Defense Policy BoardAmbassador Robert Lighthizer will chair the advisory board.

12h|651.5k4.8k763

CH#1360|@KIMMONISMUS

AC#682|@ANDREWCURRAN_

B(#530|@BEFFJEZOS

KA#403|@YACINEMTB

60

Former Meta PM Xiaoyin Qu predicts US and European enterprises will adopt Chinese AI models to self-host on private GPUsQu cites enterprise distrust of Anthropic's data retention policies

23h|483.2k2.7k1.1k

XL#1855|@XLR8HARDER

ZE#1695|@ZEPHYR_Z9

AD#1537|@DORIALEXANDER

RO#1315|@KRISHNANROHIT

RP#1257|@ROHANPAUL_AI

70

Meta restricts employees from using Claude Code and OpenAI Codex to prevent competitor model distillationThe policy protects the training pipeline of Meta's upcoming MetaCode.

17h|168.2k1.4k314

XL#1855|@XLR8HARDER

AD#1537|@DORIALEXANDER

CH#1360|@KIMMONISMUS

RP#1257|@ROHANPAUL_AI

AC#682|@ANDREWCURRAN_

B(#530|@BEFFJEZOS

80

Kim Isenberg and Peter Wang argue the US is losing the AI race due to energy and data center constraintsChina uses free models to promote domestic Huawei hardware.

21h|358k2k1k

DA#1469|@DAVIDLEE

CH#1360|@KIMMONISMUS

SB#289|@BLANCHEMINERVA

GT#121|@GARRYTAN

90

Tesla releases FSD v14 Lite, bringing distilled AI4 driving models to memory-constrained AI3 hardwareOlder AI3 computers operate with 15% of AI4's memory bandwidth

1d|#1VIEWS4.8M#1LIKED25.4k921

DM#1198|@DAVEMORIN

RS#447|@SCOBLEIZER

EM#52|@ELONMUSK

100

Smol.ai founder Shawn 'swyx' Wang opens the AI Engineer World's Fair with a focus on local AI and NVIDIA Blackwell hardwareMajor sponsors include OpenAI, Anthropic, and Google DeepMind

1d|8.6k1214

SC#1802|@MILKSANDMATCHA

BH#1762|@BENHYLAK

HH#1682|@HAMELHUSAIN

SD#1590|@STEFANIA_DRUGA

AV#1419|@ALTRYNE

PS#1404|@_PHILSCHMID

PS#1358|@PETESKOMOROCH

👩‍💻P#387|@DYNAMICWEBPAIGE

AB#228|@ABEIRAMI

110

Cognition AI co-founder Walden Yan launches Devin Fusion, a hybrid-model routing system that cuts coding agent costs by 35%It delegates targeted coding sub-tasks to a sidekick agent.

14h|43.3k317181

WA#1849|@WALDEN_YAN

AG#1848|@ITSANDREWGAO

MB#1780|@MATTHEWBERMAN

BH#1762|@BENHYLAK

EL#1136|@ELIEBAKOUCH

JL#906|@JERRYJLIU0

SW#548|@SCOTTWU46

120

Elon Musk compares machine optimization to solving an elegant puzzle that yields practical utilityResearch engineer shyamal posted in agreement with the sentiment

13h|1.6M24.6k1.1k

SH#1130|@SHYAMALANADKAT

BY#380|@BILLYUCHENLIN

EM#52|@ELONMUSK

130

Startup 8090 raises $135 million in Series A funding led by Salesforce VenturesThe capital will fund hiring and infrastructure development

18h|631.9k2.7k824

AP#1222|@APOMPLIANO

MP#928|@MARKPINC

BA#713|@ADCOCK_BRETT

CP#347|@CHAMATH

140

National Design Studio releases Rampart, a 14.7MB open-source model that redacts PII directly in the browserEdward Coristine and Tai Groot built the Hugging Face-hosted model.

13h|48.9k42878

DM#1198|@DAVEMORIN

MP#928|@MARKPINC

JC#410|@JOHNCOOGAN

LC#219|@LULUMESERVEY

C🤗#109|@CLEMENTDELANGUE

150

2025 ARC Prize winner Alexia Jolicoeur-Martineau, known for her Tiny Recursive Model work, joins MicrosoftShe joins the company from Samsung's SAIT AI Lab.

16h|54.2k1.1k79

KD#904|@CSPROFKGD

AJ#617|@JM_ALEXIA

TM#613|@ISCIENCELUVR

LB#560|@SRCHVRS

DP#217|@DIMITRISPAPAIL

JH#90|@JEREMYPHOWARD

160

CLODO launches an AI-powered talent search platform that scans 1.2 billion profiles using natural languageEarly adopters include the U.S. Intelligence Community

12h|48k10182

SA#1889|@SVPINO

SC#1802|@MILKSANDMATCHA

CH#1360|@KIMMONISMUS

RP#1257|@ROHANPAUL_AI

EL#684|@OMARSAR0

RS#447|@SCOBLEIZER

170

Claude Code creator Boris Cherny says the next version will run subagents in the background by defaultThis lets users continue chatting while tasks run independently.

13h|356.8k6.1k677

CH#1360|@KIMMONISMUS

BC#369|@BCHERNY

180

Rocket Lab to acquire Iridium Communications in an $8 billion deal, combining launch capabilities with global satellite networkRocket Lab went public in 2021 at $4.1 billion.

17h|168.8k2.2k174

PH#1772|@PETERDIAMANDIS

DE#675|@ZEBULGAR

AV#350|@ASHLEEVANCE

190

Foundation for American Innovation launches Frontier Legal Defense to fight regulatory overreach and copyright disputes in AIThe initiative is recruiting litigators fluent in AI scaling policies

16h|59.5k35376

JS#1786|@JASMINEWSUN

TH#1478|@TIMHWANG

HB#1424|@HERBIEBRADLEY

200

OpenAI teases a custom mechanical keypad designed for Codex coding shortcuts ahead of a July 15th revealThe teaser showcases a 3x3 grid and side dials.

13h|61k1k57

V(#1939|@REACH_VB

RH#642|@ROMAINHUET

TI#611|@THSOTTIAUX

210

DeepSeek's upcoming V4 models will introduce peak-hour pricing that doubles API rates during peak Beijing hoursThe models support a 1M token context window.

20h|255k1.8k204

LA#1215|@SCALING01

AC#682|@ANDREWCURRAN_

T(#501|@TEORTAXESTEX

220

OpenAI's Gabriel Petersson argues strategic pivots and intense, founder-type hires make AI companies structurally prone to internal conflictCreator roon countered by highlighting Anthropic's internal cohesion.

17h|108.3k994153

YI#1023|@YISHAN

J⧉#674|@REPLIGATE

GA#598|@GABRIEL1

230

Jacob Andreas, MIT NLP research lead, co-develops method to replace transformer attention heads with human-readable Python codeReplacing 25% of heads maintained downstream question-answering accuracy.

15h|23.1k14279

LR#897|@LAURARUIS

BL#764|@BELINDAZLI

RS#741|@ZIV_RAVID

SR#336|@SIVAREDDYG

JA#36|@JACOBANDREAS

240

Meta's Lucas Beyer asks if any serious open-source LLM pretraining repositories exist without relying on NVIDIA's Megatron-LMAndrew Carr proposed PyTorch's torchtitan as a potential alternative

11h|17.9k18487

EL#1136|@ELIEBAKOUCH

AC#484|@ANDREW_N_CARR

SB#289|@BLANCHEMINERVA

LB#72|@GIFFMANA

250

Lindy and Cursor shift production workloads to Chinese models DeepSeek v4 and Kimi K2Shopify and Airbnb are also reportedly integrating Chinese models.

10h|32.1k30465

BR#1861|@BRIANROEMMELE

DJ#1706|@DAN_JEFFRIES1

KC#552|@KYLECRANMER

KA#403|@YACINEMTB

260

Anthropic renegotiates Amazon cloud deal, shifting to a token-based structure that increases costs for AmazonThe structure could increase downstream costs for end users.

14h|98.1k923114

LA#1215|@SCALING01

JL#1086|@JESSICALESSIN

AC#682|@ANDREWCURRAN_

270

Groq founder Jonathan Ross argues startups should hire interns because LLMs let early-career talent bypass conventional technical limitationsMarkus Wulfmeier noted student-led projects drove top DeepMind research.

17h|77.5k497125

FB#1778|@XEOPHON

JR#1280|@JONATHANROSS321

MW#685|@M_WULFMEIER

NL#80|@NATOLAMBERT

280

Three labor economists debate whether AI will reverse the Industrial Revolution by making human labor abundantDavid Autor argues AI will create new human tasks.

21h|25.6k181151

RP#1257|@ROHANPAUL_AI

RI#232|@RISHIBOMMASANI

290

TheStalwart newsletter argues geopolitical AI competition could force governments to use domestic austerity to fund computing powerCreator @teortaxesTex says resource diversion already drives inflation concerns

14h|143.6k546267

DF#756|@DANIELLEFONG

T(#501|@TEORTAXESTEX

300

Benchmark's Everett Randle discusses the state of the startup ecosystem in a Sourcery interviewThe 57-minute interview contains no product or technical demonstrations.

16h|261.5k379285

PC#1025|@PARKERCONRAD