Tech News

Today's Highlights

In case you missed it

SpaceX agrees to acquire Anysphere, creator of the AI code editor Cursor, in a $60 billion all-stock merger

DE#274|@DEEDYDAS

GM#178|@GARYMARCUS

EM#171|@EMOSTAQUE

EM#52|@ELONMUSK

#1 VIEWED

OpenAI Codex engineering lead Thibault Sottiaux says Codex tools can now be configured to run any open-source model

TI#611|@THSOTTIAUX

TW#27|@THOM_WOLF

RISING LIKES

282

Z.ai's GLM-5.2 Max takes third on Artificial Analysis Index, delivering 112 tokens per second for $0.41 per task

CH#1360|@KIMMONISMUS

T(#501|@TEORTAXESTEX

FASTEST CLIMBING

DeepSeek's Deli Chen open-sources SKILL, an autonomous ML research framework that executed a 285-billion parameter GRPO experiment

DC#1803|@VICTOR207755822

T(#501|@TEORTAXESTEX

Top Stories

Posts:—|Clusters:—|Next Crawl:—

10

CEOs of Anthropic, OpenAI, DeepMind, and Mistral meet at G7 amid Trump administration dispute over model accessThe dispute centers on models with elite cyber capabilities.

3h|173.6k3.2k317

ZE#1695|@ZEPHYR_Z9

TH#1478|@TIMHWANG

CH#1360|@KIMMONISMUS

ME#861|@MERVENOYANN

RS#741|@ZIV_RAVID

JC#335|@JULIEN_C

Top comment: @julien_c “@etnshow”

20

OpenAI Codex engineering lead Thibault Sottiaux says Codex tools can now be configured to run any open-source modelUsers can route client traffic to self-hosted local endpoints.

5h|#1VIEWS442.7k#1LIKED4k#1BOOKMARKED1.8k#1COMMENTS534

TI#611|@THSOTTIAUX

TW#27|@THOM_WOLF

Top comment: @FilipBaturan “@thsottiaux Hey Tibo, is it possible to set up a hybrid approach in Codex where we use GPT models as architects and open source models as executors? I'm trying to get it working, but it seems Codex expects the same tool-calling protocol from other models, which they don’t have.”

31

Next.js creator Guillermo Rauch's Vercel launches Eve, a file-based AI agent framework with a unified Agent StackIt uses Markdown for instructions and TypeScript for tools.

6h|110.3k1.5k640

EL#684|@OMARSAR0

Top comment: @rauchg “@kumard_3 Yes. In a very tangible way I can describe in this tweet. Deploying eve basically gives you an API backend. And then you can query it with a 𝚞𝚜𝚎𝙴𝚟𝚎𝙰𝚐𝚎𝚗𝚝() hook we produce, for Next, Vite, TanStack to consume. https://eve.dev/docs/guides/frontend/overview”

Watchlist

Video Signals

CH#1360|@KIMMONISMUS

Trump Joins AI CEOs Altman And Amodei At G7 Summit Lunch In France

53m ago|Views 13KLikes 381Bookmarks 91

Vercel Unveils New Homepage Spotlighting Agentic Infrastructure

6h ago|Views 84KLikes 698Bookmarks 155

Saylor Explores Bitcoin AI And Digital Finance Strategies At BTC Prague

2h ago|Views 93KLikes 2.2KBookmarks 380

SA#1889|@SVPINO

MaineCoon Unveils First Unlimited-Duration Real-Time Interactive Audio-Visual AI Model

3h ago|Views 5.5KLikes 34Bookmarks 18

DD#1499|@DIMADAMEN

HOI-DETR Detects Hands Objects And Interactions From Single Images

3h ago|Views 790Likes 28Bookmarks 5

Graham Neubig Releases Agent Canvas Frontend For Building AI Agents

1h ago|Views 1.6KLikes 26Bookmarks 9

EM#52|@ELONMUSK

Elon Musk Shares Cybertruck Chase Video Made With Grok Imagine 1.5

6h ago|Views 13KLikes 0Bookmarks 0

CP#347|@CHAMATH

Chamath Palihapitiya Spotlights Elite AI Engineer Sai at 8090 Factory

29m ago|Views 13KLikes 75Bookmarks 24

EL#684|@OMARSAR0

DAIR AI Academy Showcases Interactive Neural Network Lessons With Code Editor

11m ago|Views 166Likes 4Bookmarks 4

JZ#753|@JAMES_Y_ZOU

Oxford And Stanford Launch Data2Story Agent For Verifiable Multimodal Data Stories

1h ago|Views 2.6KLikes 0Bookmarks 0

41

Europe 2031 initiative warns three individual US AI labs each control more compute than all of Europe combinedMIT's Michiel Bakker backed the warning about Europe's trajectory.

5h|7.2k8513

MB#1589|@BAKKERMICHIEL

MD#1519|@MICHAELD1729

AK#271|@BLACKHC

Top comment: @bakkermichiel “Scenario: https://europe2031.ai Summary: https://europe2031.ai/summary Audio: https://europe2031.ai/audio/ The ⭐⭐⭐ team: Daan Juijn, Stan van Baarsen @DadaJudith Lily Stelling, @philip_fox_ @AlexTPet @TomChivers”

51

Gavin Baker, Atreides Management CIO, argues AI revenue per gigawatt can rise if token throughput gains outpace falling token pricesDavid Lee compares falling token prices to Google's CPC decline.

2h|43k394158

DA#1469|@DAVIDLEE

GB#449|@GAVINSBAKER

Top comment: @GavinSBaker “Original post:”

New github stars (48hrs)

Terminal AI coding assistant that inspects codebases edits files and runs multi-step workflows

Noumena-Network/code3513h

Provides terminal AI coding assistant for codebase inspection, file edits, command execution and multi-step workflows.

Hardware-isolated sandbox with allowlist firewall and monitor for safe Claude Code sessions

alexander-turner/claude-guard2617h

Wraps Claude Code sessions with deterministic checks on commands and writes to limit agent access.

61

GLM-5.2 reintroduces critic and value models to reinforcement learning, moving away from GRPO for long-horizon tasksJeremy Howard highlighted the limitations of group-based variance reduction.

5h|3.2k665

XL#1855|@XLR8HARDER

T(#501|@TEORTAXESTEX

JH#90|@JEREMYPHOWARD

Top comment: @hallerite “It's the same principle behind ECHO (SFTing on tool-call outputs), if you understand RL in an information theoretic way. We simply need to get as much out of every rollout as possible. Naive GRPO is unbelievably wasteful.”

71

Stanford NLP postdoc Peter Hase joins Schmidt Sciences to fund scientific research on AI systemsHase will maintain his Stanford affiliation for hands-on research

1h|3.5k985

PH#836|@PETERBHASE

MB#208|@MOHITBAN47

Top comment: @mohitban47 “@peterbhase Congrats again @peterbhase on the new Program Lead role at Schmidt Sciences -- proud of you for creating these very timely and important grant opportunities for the wider academic/research community!! #ProudAdvisor 🤗”

81

Z.ai's GLM-5.2 Max takes third on Artificial Analysis Index, delivering 112 tokens per second for $0.41 per taskThe standard GLM-5.2 variant scored ahead of Gemini 3.5 Flash.

2h|14.7k28216

RISING LIKES282

CH#1360|@KIMMONISMUS

T(#501|@TEORTAXESTEX

Top comment: @kimmonismus “https://artificialanalysis.ai/models/glm-5-2”

92

AI2 post-training lead Nathan Lambert argues that the core efficacy of both PPO and GRPO relies on policy gradient methodsHe authors the RLHF Book and Interconnects.ai newsletter.

1h|9.9k16537

XL#1855|@XLR8HARDER

NL#80|@NATOLAMBERT

Top comment: @MRashadnow “@natolambert People will read an experiment from an ai lab and start announcing the death of stuff”

108

DeepSeek's Deli Chen open-sources SKILL, an autonomous ML research framework that executed a 285-billion parameter GRPO experimentThe autonomous agent wrote its own paper, scoring 8.6.

28m|7.1k5245

FASTEST CLIMBING+8

DC#1803|@VICTOR207755822

T(#501|@TEORTAXESTEX

Top comment: @teortaxesTex “@victor207755822 thanks! Deli, what do you (or your agents) think of Zhipu's argument about PPO being strictly preferable for long multi-turn agentic RL?”

112

Wayve, Stellantis, and Uber partner to build and deploy Level 4 robotaxis using end-to-end deep learningUber will deploy the completed vehicles on its network.

7h|22.3k1083

AK#1652|@ALEXGKENDALL

Top comment: @AlexRoy144 “@alexgkendall @wayve_ai @Stellantis @Uber This will be big.”

120

Google Cloud launches Open Knowledge Format to standardize LLM-wiki context structures for AI agentsThe portable format uses markdown files with YAML frontmatter.

7h|4.5k3024

CB#1674|@CODE_STAR

AK#271|@BLACKHC

Top comment: @tekbog “@code_star it feels just like a bridge because we don’t have anything better tbh”

131

AI commentator Teortaxes concedes incorrect prediction as GLM-5 scores 40 on the Artificial Analysis indexTeortaxes previously predicted GLM 5.2 would score 48.

1h|7.2k1237

T(#501|@TEORTAXESTEX

KA#403|@YACINEMTB

Top comment: @yacineMTB “@teortaxesTex I am so happy for techshrek”

141

Jennifer finds changing her name to Jeff on an AI resume tool triggers more limiting career feedbackGary Marcus highlighted the financial scale of the biased tool.

8h|6.4k487

GM#178|@GARYMARCUS

Top comment: @abowenwk “@GaryMarcus AI is misaligned because it's not able to differentiate itself. That's a big issue for psychology. https://abowenkc.substack.com/p/the-devolution-of-society-differentiation?r=72tmbx”

158

Neko Health's £299 preventive full-body health scan bundles heart, skin, and blood tests into a one-hour London appointmentReddit co-founder Alexis Ohanian called for similar US options

4h|7k351

DE#638|@ELDSJAL

AO#636|@ALEXISOHANIAN

Top comment: @alexisohanian “@mignano @Neko Preach!! Thanks for trying & sharing.”

161

Systems engineer Yacine argues RL for robotics and Atari builds better foundations than working on LLMs directlyJoseph Suarez agreed, calling RL for LLMs overcomplicated and bloated.

1h|3k568

JS#1687|@JSUAREZ

KA#403|@YACINEMTB

Top comment: @allisonology “@yacineMTB Plausible; when is your next space yacine any new coffees you've had recently?”

171

Florian Brand, who works on LLM evaluations at Prime Intellect, teases unresolved upstream bugs in popular AI benchmarks without detailsThe upstream benchmark issues have persisted for over a year.

4h|2.3k836

XL#1855|@XLR8HARDER

FB#1778|@XEOPHON

Top comment: @xeophon “@xlr8harder no kink shaming in my comments pls”

181

Arcee AI's Cody Blakeney says LLMs increasingly fabricate novel hyphenated compound adjectives instead of overusing em dashesThe pattern makes LLM-generated text stylistically more German.

7h|2.5k350

FB#1778|@XEOPHON

CB#1674|@CODE_STAR

Top comment: @YannickScholich “@code_star @xeophon Lieben wir, cheers mate here another beauty: Gärungsgrundlagenbereitstellungsprozessoptimierungsbeauftragtenfortbildungsrichtlinie”

191

LisanBench creator Lisan al Gaib argues prompt engineering is becoming obsolete as LLMs learn to infer user intent automaticallyCurrent models already demonstrate adaptive prompting and identity inference.

7h|1.2k162

XL#1855|@XLR8HARDER

LA#1215|@SCALING01

Top comment: @MoonL88537 “@scaling01 @aidan_mclau >they are inferring intent and who they are talking to a model *literally could not work* if this wasn't true. i vote no: asi will understand intent so deeply that 'prompting' is no longer a thing.”

201

Hugging Face founder Clement Delangue says users underestimate AI hosting costs and resist paying $9 monthly feesYacine noted high S3 bills are shutting down competitors.

1h|1.4k200

KA#403|@YACINEMTB

C🤗#109|@CLEMENTDELANGUE

Top comment: @ClementDelangue “@yacineMTB yeah, not enough people understand how hard it is to do what the HF team does at scale (and start complaining as soon as we ask them to pay $9 a month haha)”

Rising Stories

German Word Test7H AGO
Claude Needs 31-Letter German Word for True AGI
#1674
5257
Staged Scaling Push17M AGO
Staged Training Needed for 10T Models
#1136
265

Recent Stars

…irschstein/flexavatar2H AGO
Learns complete 3D head avatars from partial image supervision using neural rendering.
484
34 stars
zai-org/SCAIL-211H AGO
Provides inference code and tools for SCAIL-2, enabling end-to-end character animation and replacement via unified motion conditioning.
484
569 stars
Noumena-Network/code13H AGO
Provides terminal AI coding assistant for codebase inspection, file edits, command execution and multi-step workflows.
573
1136
35 stars
langchain-ai/openevals15H AGO
Provides prebuilt evaluators, LLM-as-judge prompts, and frameworks for assessing LLM app outputs including extraction and tool calls.
1014
1.1k stars
…ngchain-ai/agentevals15H AGO
Supplies evaluators for agent trajectories via exact matching modes and LLM judges.
1014
622 stars

Github Stars

(7 days)

macrodata-labs/refiner2D AGO
Processes and refines large-scale ML datasets via a data framework.
1719
1136
353
888
1674
51 stars
Noumena-Network/code13H AGO
Provides terminal AI coding assistant for codebase inspection, file edits, command execution and multi-step workflows.
573
1136
35 stars
…r-turner/claude-guard17H AGO
Wraps Claude Code sessions with deterministic checks on commands and writes to limit agent access.
1278
570
26 stars
omnigent-ai/omnigent3D AGO
Implements a domain-agnostic framework for autonomous AI agents with ReAct loops, multi-LLM routing, reasoning graphs and MCP integration.
586
1066
1.6k stars

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

Yesterday's Top Stories, Jun 16, 2026.

Posts: 4,163|Clusters: 1,400|Stories frozen at midnight PT

10

SpaceX agrees to acquire Anysphere, creator of the AI code editor Cursor, in a $60 billion all-stock mergerThey recently co-trained a new model for Grok Build.

19h|20.2M98.5k7.4k

FB#1778|@XEOPHON

DM#1766|@DAVEMCCLURE

BH#1762|@BENHYLAK

RL#1733|@RYOLU_

ND#1653|@NICKADOBOS

B(#1623|@ANDERSONBCDEFG

CB#1428|@CHRISJBAKKE

PS#1404|@_PHILSCHMID

CP#1392|@CHETANP

SI#1378|@SUZATWEET

CH#1360|@KIMMONISMUS

SO#1349|@SOLEIO

RP#1257|@ROHANPAUL_AI

JY#1244|@JYANGBALLIN

SM#1225|@PITDESI

KR#1220|@KYLIEBYTES

LA#1215|@SCALING01

TN#1157|@TURNERNOVAK

DZ#1153|@ZDHNARSIL

EL#1136|@ELIEBAKOUCH

SH#1130|@SHYAMALANADKAT

SW#1092|@SJWHITMORE

JL#906|@JERRYJLIU0

KD#904|@CSPROFKGD

NM#793|@MUENNIGHOFF

DF#756|@DANIELLEFONG

EL#684|@OMARSAR0

@L#627|@LEVELSIO

TM#613|@ISCIENCELUVR

YJ#586|@YUCHENJ_UW

SM#550|@SHAUNMMAGUIRE

B(#530|@BEFFJEZOS

AS#478|@AMANRSANGER

RP#395|@TYPESFAST

BP#381|@BORISMPOWER

BY#380|@BILLYUCHENLIN

CP#347|@CHAMATH

MT#346|@MNTRUELL

DE#274|@DEEDYDAS

GM#178|@GARYMARCUS

EM#171|@EMOSTAQUE

EM#52|@ELONMUSK

20

Z.ai releases open-weights GLM-5.2 with a 1-million-token context window, beating Claude Opus 4.7 on Code Arena FrontendThe MIT-licensed model features two cost-balancing reasoning modes.

3d|1.5M13.8k2.3k

BR#1879|@BINDUREDDY

VV#1864|@VIPULVED

WH#1824|@NREHIEW_

AS#1809|@ALEXSTAMOS

AJ#1798|@ALEXJC

FB#1778|@XEOPHON

BZ#1718|@BANGHUAZ

ZE#1695|@ZEPHYR_Z9

CB#1674|@CODE_STAR

AV#1419|@ALTRYNE

CH#1360|@KIMMONISMUS

LA#1215|@SCALING01

MA#1207|@MAKSYM_ANDR

YP#1171|@YPATIL125

AN#1147|@ML_ANGELOPOULOS

EL#1136|@ELIEBAKOUCH

VW#990|@VINCENTWEISSER

CL#942|@CLATTNER_LLVM

ME#861|@MERVENOYANN

ZW#838|@ZHAORAN_WANG

DF#756|@DANIELLEFONG

RS#741|@ZIV_RAVID

EL#684|@OMARSAR0

AC#682|@ANDREWCURRAN_

TM#613|@ISCIENCELUVR

T(#501|@TEORTAXESTEX

AC#484|@ANDREW_N_CARR

RS#447|@SCOBLEIZER

KE#432|@ETHAYARAJH

KA#403|@YACINEMTB

AD#166|@ALEXGDIMAKIS

GT#121|@GARRYTAN

C🤗#109|@CLEMENTDELANGUE

NL#80|@NATOLAMBERT

30

Trump administration orders Anthropic to suspend global access to Fable 5 and Mythos 5 under a 90-minute deadlineForeign nationals are now barred from training these models.

4d|916k4.9k992

BR#1879|@BINDUREDDY

ZE#1695|@ZEPHYR_Z9

TH#1478|@TIMHWANG

JL#1425|@JEFFLADISH

CH#1360|@KIMMONISMUS

RO#1315|@KRISHNANROHIT

RP#1257|@ROHANPAUL_AI

LA#1215|@SCALING01

CA#1059|@CHRISALBON

DF#756|@DANIELLEFONG

RS#741|@ZIV_RAVID

AC#682|@ANDREWCURRAN_

C(#567|@STEPHENLCASPER

T(#501|@TEORTAXESTEX

KA#403|@YACINEMTB

AK#271|@BLACKHC

DR#113|@DELIPRAO

LB#72|@GIFFMANA

40

NVIDIA's GEAR Lab launches ENPIRE for autonomous robot policy self-improvement in the real worldEight parallel agents autonomously learned to install physical GPUs.

14h|496.7k3.4k1.6k

TS#1677|@TIMSHI_AI

GS#1558|@GUANYASHI

LA#1215|@SCALING01

KG#996|@KEERTHANPG

SL#995|@SHUANGL13799063

CP#855|@CHRIS_J_PAXTON

SZ#785|@STEPHZHAN

MM#716|@MASCOBOT

KG#508|@KEN_GOLDBERG

KA#403|@YACINEMTB

YS#397|@YSU_NLP

JF#66|@DRJIMFAN

50

Elon Musk proposes capturing space-based solar energy to scale AI computation beyond human capacityThe theoretical roadmap leverages resources from Tesla, SpaceX, and xAI

16h|#1VIEWS25.1M#1LIKED190.3k#1BOOKMARKED9.3k

TN#1157|@TURNERNOVAK

NR#502|@NAVEENGRAO

EM#52|@ELONMUSK

60

Cursor announces Origin, a Git-compatible, MCP-extensible code hosting platform built for high-throughput AI agentsThe hybrid NVMe and S3 architecture supports infinite replicas.

13h|763k5.4k1k

CR#1942|@CWOLFERESEARCH

BR#1861|@BRIANROEMMELE

ND#1653|@NICKADOBOS

AV#1419|@ALTRYNE

AV#1193|@AVISCHIFFMANN

SW#1092|@SJWHITMORE

B(#530|@BEFFJEZOS

70

Microsoft shifts Copilot Cowork to usage-based pricing and tests self-hosted DeepSeek V4 to reduce infrastructure costsFlat-rate pricing became unsustainable due to high-volume users.

13h|909.5k5.4k900

ZE#1695|@ZEPHYR_Z9

PS#1594|@RECURSEPARADOX

CH#1360|@KIMMONISMUS

RO#1315|@KRISHNANROHIT

AC#682|@ANDREWCURRAN_

GM#178|@GARYMARCUS

80

Voice AI startup Bland raises $100 million Series C to automate complex, regulated telephony workflowsA comedic launch video featured founder Paul Graham.

13h|1M4.7k2.6k

MS#1418|@MSUSTER

AC#1371|@ANOTHERCOHEN

RP#1257|@ROHANPAUL_AI

MA#1240|@ARRINGTON

GT#121|@GARRYTAN

AK#33|@_AKHALIQ

90

Andreessen Horowitz releases a custom video tribute tracing SpaceX's trajectory from early failures to aerospace milestonesThe video opens by disclosing a16z's SpaceX shareholder status.

1d|12.3M71.9k6.3k

ET#524|@ERIKTORENBERG

EM#52|@ELONMUSK

100

Cursor previews an upcoming 1.5-trillion-parameter custom LLM trained from scratch on more than 100,000 GPUsThe model targets general intelligence and launches within weeks.

13h|642.6k7.8k1.1k

ND#1653|@NICKADOBOS

LA#1215|@SCALING01

TA#1140|@VICTORTAELIN

AC#682|@ANDREWCURRAN_

B(#530|@BEFFJEZOS

110

xAI releases Grok Imagine Video 1.5, including a fast model that generates 720p video in 25 secondsIt includes multi-agent tools for consistent multi-shot narrative sequences.

2h|1.2M7.6k467

BR#1861|@BRIANROEMMELE

ZH#1073|@ZHITINGHU

RS#447|@SCOBLEIZER

EM#52|@ELONMUSK

120

OpenAI resolves Codex capacity issues by resetting model rate limits over a 24-hour windowUsers previously encountered "Selected Model is at Capacity" errors

18h|1.4M10.4k545

V(#1939|@REACH_VB

MB#1780|@MATTHEWBERMAN

TI#611|@THSOTTIAUX

130

Coinbase CEO Brian Armstrong announces plans for tokenized U.S. stocks backed 1:1 by actual sharesSmart contracts will automatically distribute dividend payments to holders

16h|504k5.1k348

EC#1632|@EMILIEMC

AG#690|@ANTONIOGM

BA#398|@BRIAN_ARMSTRONG

140

Anthropic says Claude Code success rates rise from 77% to 91% as user domain expertise increasesCritics question the study's reliance on Claude self-evaluation

10h|140.7k1.6k479

CR#1942|@CWOLFERESEARCH

FB#1778|@XEOPHON

LA#1215|@SCALING01

RS#741|@ZIV_RAVID

AC#682|@ANDREWCURRAN_

DY#120|@DIYI_YANG

Digg Deeper question: “What type of instructions did expert developers give that set them apart from novices?”

150

Gergely Orosz reports Meta is walking back harsh engineering directives, including 24/7 screen recording and reassignment to AI data labelingReassigned engineers solved LeetCode to train coding AI models

23h|1M12.1k1.3k

GO#1155|@GERGELYOROSZ

DA#619|@GROWING_DANIEL

GM#178|@GARYMARCUS

160

Elon Musk shares a Grok Imagine video generation demo, which investor Steve Jurvetson notes resembles a physical artwork he ownsThe clip showcases detailed rendering of fluids and light.

16h|3.8M27.7k1.3k

SJ#1274|@FUTUREJURVETSON

EM#52|@ELONMUSK

170

Thais Branco's Taste Labs emerges from stealth with $18.5 million to build data infrastructure that eliminates 'AI slopInitial applications focus on design and model evaluation.

14h|123.5k651290

BH#1762|@BENHYLAK

TS#1677|@TIMSHI_AI

GR#1176|@GOKULR

RK#1011|@MIKEISAAC

NR#985|@NAZNEENRAJANI

SC#786|@SARAHCAT21

KM#409|@KORYMATH

SD#104|@_SHOLTODOUGLAS

180

Coinbase registers its Coinbase Advisor AI agent with the SEC to offer personalized financial adviceThe tool rolls out first to Coinbase One subscribers.

11h|528.8k3.4k917

MS#1765|@MATTPRD

EC#1632|@EMILIEMC

BA#398|@BRIAN_ARMSTRONG

190

OpenAI rolls out Codex Computer Use, Chrome integration, and Memory to EU and UK users, powered by GPT-5.3-Codex-SparkThe agent controls local Mac applications by capturing screenshots

14h|385.4k2.8k964

V(#1939|@REACH_VB

TI#611|@THSOTTIAUX

200

SpaceX acquires Cursor, sparking debate over the coding assistant's strategy and its origins as a mechanical CAD toolThrive Capital backed founder Michael Truell before the pivot.

18h|262.4k3.3k387

BH#1762|@BENHYLAK

210

Apple plans to launch AirPods with built-in cameras in late 2027 to power visual AI featuresThe cameras are for visual intelligence, not capturing photos.

14h|1.5M5.2k592

RS#447|@SCOBLEIZER

220

Leaked audited financials show OpenAI made $13.07 billion in revenue against $34 billion in costs in 2025SoftBank and Microsoft contributed $1.17 billion of that revenue.

21h|143.8k785275

CH#1360|@KIMMONISMUS

RP#1257|@ROHANPAUL_AI

MS#585|@MARTINSHKRELI

AK#271|@BLACKHC

GM#178|@GARYMARCUS

Digg Deeper question: “How does this compare the anthropics revenue/losses?”

230

DeepSeek raises $7.4 billion at a $50 billion valuation with terms that secure absolute control for founder Liang WenfengWenfeng personally contributed $3 billion of the total capital

1d|267k2.1k273

ZE#1695|@ZEPHYR_Z9

RP#1257|@ROHANPAUL_AI

TM#613|@ISCIENCELUVR

T(#501|@TEORTAXESTEX

240

OpenAI releases "Deployment Simulation" methodology, claiming 92% directional accuracy in predicting model safety risks before launchIt replays de-identified conversation logs to test unreleased models.

11h|21.2k280106

B(#1623|@ANDERSONBCDEFG

TK#1513|@TOMEKKORBAK

HS#1480|@HANNAHSHEAHAN

RP#1257|@ROHANPAUL_AI

MC#936|@MICAHCARROLL

LL#521|@LUCY3_LI

MB#57|@MILES_BRUNDAGE

250

Ethan Mollick says open-source models lag closed-source by a year, but investor Ramez Naam argues the gap is only four monthsMollick warns firms have months to harden IT systems.

13h|140.8k1.3k181

B(#1623|@ANDERSONBCDEFG

T(#501|@TEORTAXESTEX

KA#403|@YACINEMTB

EM#184|@EMOLLICK

260

San Francisco media startup Nell launches with backing from Index Ventures, Soleio Cuervo, and Guillermo RauchSpecific product features and technical details remain undisclosed.

14h|383.1k2.8k2k

SO#1349|@SOLEIO

LR#543|@LENNYSAN

270

The Boring Company completes commissioning of its Prufrock-MB2 tunneling machine in Nashville after an 11 rpm testThe company plans to ship its Prufrock-MB3 machine in August

3h|1.4M18.8k501

B(#530|@BEFFJEZOS

EM#52|@ELONMUSK

280

Genesis announces Eno, its first general-purpose robot, featuring a non-humanoid design slated for a Q4 releaseObservers praised its physical fluidity compared to Tesla's Optimus

17h|54.8k37557

SA#1889|@SVPINO

BR#1861|@BRIANROEMMELE

SH#1443|@ZERGYLORD

KD#904|@CSPROFKGD

CP#855|@CHRIS_J_PAXTON

DC#807|@DCHAPLOT

B(#530|@BEFFJEZOS

VK#161|@VKHOSLA

290

Mistral AI CEO Arthur Mensch announces a new family of large but sparse models debuting this summerRestricted early access begins in July for key partners.

14h|304.2k2.3k128

LA#1215|@SCALING01

EL#1136|@ELIEBAKOUCH

DF#756|@DANIELLEFONG

T(#501|@TEORTAXESTEX

AM#256|@ARTHURMENSCH

300

Snap CEO Evan Spiegel debuts bulky new Specs AR glasses, sending shares down following the public revealCritics compared the bulky frames to 3D cinema glasses.

6h|118.8k1.8k87

AC#1371|@ANOTHERCOHEN

AL#1154|@ALTH0U

DA#619|@GROWING_DANIEL