Tech News

Today's Highlights

Ramp and Revelio Labs study of 21,559 firms finds high-intensity AI adopters grew headcount by 10.2%

+6 more

It’s a new day at Digg. Fresh stories are clustering now. 50% complete.

Top Stories

Posts:—Clusters:—Next Crawl:—

Meta plans to launch a cloud service selling excess AI compute capacity and hosted models to offset infrastructure costsThe service will compete with AWS Bedrock and CoreWeave.

5h32.4k18312

#178

Top comment: @kimmonismus “source: https://www.bloomberg.com/news/articles/2026-06-30/us-government-lifts-restrictions-on-anthropic-s-fable-5-model?srnd=homepage-europe”

Founder @levelsio warns the EU's new sub-€150 customs duty will trigger severe shipping delays and complex tax burdensRecipients must manually classify imports using complex tariff codes.

6h6.4k949

#395

Top post: @levelsio “As someone who lives in Portugal which already stops every single package you get from outside EU with no exceptions for a decade or more: I can tell you how this will go From now on if you receive a package from outside EU, it'll first go missing for weeks and nobody will know where it is, then you'll receive an email from your country's customs agency They'll make you login to some extremely unusable and outdated web portal which breaks half the time and where you have to classify your shipment in one of 10,000 confusing category codes Depending on the category code you'll pay a different tax rate, sometimes it's 5% sometimes it's 60% Then you have to pay the customs duty in some archaic way that will also be broken half the time And then you'll finally receive your package from outside EU after months Experiencing this for a few times you'll realize you will never ever try to order or receive anything from outside EU ever because the ordeal is too painful and expensive I get asked a lot here by people from US to test out new products (like the guy who made the Minimal Phone) and it's just literally impossible for me to receive it even for testing. I tried to classify it as samples for testing but you need to go through a bureaucratic process of insane lengths to get that through so I just gave up Realistically all this does is make Europeans live even more in the stone ages, as now they can't use modern AI, nor get modern products from outside EU because it'll be too much hassle to order”

Google DeepMind's Pranav Shyam argues short-horizon PPO reduces to bandit problems, but open-source developer Grad disputes the latency impactGrad says 20,000-token reasoning workloads bypass critic bottlenecks.

9h15162

#843

Top comment: @Grad62304977 “Like my point is for long horizon I would trust the value models less and less and with compute scaling I would prefer a bigger group size or other ideas to improve credit assignment here If we assume the value model is pretty good then surely it should’ve always been what we should’ve been doing and save lots of compute still”

Watchlist

Video Signals

#1571

AI Agents Trigger Rising Privacy Concerns As Shared Dependencies Shift

4h ago28610

#1857

AI Subagent Scripts 11 Hand Transformations Using Omni and Lyria 3

10h ago850104

#1260

Perplexity CEO: Power Users Now Consume Millions in AI Compute

10h ago97300

#1311

Robot Juggling Skills Progress From Three To Five Balls In Two Attempts

9h ago1.1K00

#1893

SuperNori Debuts Proactive AI Agent For Family Caretakers

11h ago2.2K00

#856

Open Source DexSkin Adds Deformable Touch Sensing To Robot Hands

4h ago12410

T-Mobile SuperBroadband Demo Keeps AI Running After Connectivity Cut

5h ago96610

Physical AI Models Simulate Real-World Chaos for Robotics Training

7h ago56331

UBTECH Walker S2 Humanoid Robot Rallies with Human in Live Tennis Match

5h ago29900

Wireless EMG Bionic Hand Delivers Independent Finger Control

8h ago25300

US House advances the KIDS Act to regulate AI chatbots and mandate web age verificationCritics warn the rules could force universal web identity verification.

8h65970

#501

Top comment: @RepThomasMassie “Here’s the roll call: https://clerk.house.gov/evs/2026/roll228.xml”

Meituan open-sources LongCat-2.0, a 1.6-trillion-parameter MoE coding model trained on 50,000 Ascend 910 chipsThe model features a 1-million-token context window.

9h704172

#1260

Top comment: @rohanpaul_ai “https://www.longcatai.org/”

AI chip startup Etched raises $800 million and tapes out its chip after deploying an emergency sprint to BangaloreThe company has secured over $1 billion in customer contracts.

6h1.1k00

#571

Top comment: @CMeswani “@Tijori1 They are literally saying that the Bangalore team (vendor) was putting them behind schedule and that's why they had to stay here. That's not a pat on the back.”

Technical commentators joke that a highly recursive post by Paul Graham resembles Lisp without parenthesesCommentators joked the nested sentence structure needed closing parentheses

8h53970

#797

Top post: @bscholl “@paulg Needs more parentheses”

Prime Intellect's kalomaze argues GANs remain ubiquitous because they power the autoencoders used in modern diffusion modelsThe debate counters claims that the technology is obsolete.

9h23530

#598

Top post: @gabriel1 “@willdepue GANs are hella satisfying”

Elie Bakouch says Claude's safety filters blocked a routine request to retrieve an Artifact link from a past sessionThe filter flagged the cross-session request as a policy violation

10h10020

#1138

Top post: @eliebakouch “@xeophon you know the worst part, the color was like blue for claude and orange for codex and just asked in another convo to switch it and send the artifact link 😭”

100

Google DeepMind's Pranav Shyam defines short context horizons at 32k tokens or fewer and long context at 64k or moreLonger horizons require complex multi-round tool calls and code execution.

9h1820

#1525

Top comment: @Grad62304977 “@recurseparadox but whats the intuition here that tool calls and environments have a big effect here. As in its a clear point the value model can make a more accurate prediction on the final expected reward? Why is it different to say a long reasoning chain with a clear step made?”

Recent Stars

Github Stars

(7 days)

Today's Highlights

Ramp and Revelio Labs study of 21,559 firms finds high-intensity AI adopters grew headcount by 10.2%

Top Stories

Meta plans to launch a cloud service selling excess AI compute capacity and hosted models to offset infrastructure costsThe service will compete with AWS Bedrock and CoreWeave.

Founder @levelsio warns the EU's new sub-€150 customs duty will trigger severe shipping delays and complex tax burdensRecipients must manually classify imports using complex tariff codes.

Google DeepMind's Pranav Shyam argues short-horizon PPO reduces to bandit problems, but open-source developer Grad disputes the latency impactGrad says 20,000-token reasoning workloads bypass critic bottlenecks.

Video Signals

US House advances the KIDS Act to regulate AI chatbots and mandate web age verificationCritics warn the rules could force universal web identity verification.

Meituan open-sources LongCat-2.0, a 1.6-trillion-parameter MoE coding model trained on 50,000 Ascend 910 chipsThe model features a 1-million-token context window.

AI chip startup Etched raises $800 million and tapes out its chip after deploying an emergency sprint to BangaloreThe company has secured over $1 billion in customer contracts.

Technical commentators joke that a highly recursive post by Paul Graham resembles Lisp without parenthesesCommentators joked the nested sentence structure needed closing parentheses

Prime Intellect's kalomaze argues GANs remain ubiquitous because they power the autoencoders used in modern diffusion modelsThe debate counters claims that the technology is obsolete.

Elie Bakouch says Claude's safety filters blocked a routine request to retrieve an Artifact link from a past sessionThe filter flagged the cross-session request as a policy violation

Google DeepMind's Pranav Shyam defines short context horizons at 32k tokens or fewer and long context at 64k or moreLonger horizons require complex multi-round tool calls and code execution.

Recent Stars

Github Stars

Today's Highlights

Ramp and Revelio Labs study of 21,559 firms finds high-intensity AI adopters grew headcount by 10.2%

Top Stories

Meta plans to launch a cloud service selling excess AI compute capacity and hosted models to offset infrastructure costsThe service will compete with AWS Bedrock and CoreWeave.

Founder @levelsio warns the EU's new sub-€150 customs duty will trigger severe shipping delays and complex tax burdensRecipients must manually classify imports using complex tariff codes.

Google DeepMind's Pranav Shyam argues short-horizon PPO reduces to bandit problems, but open-source developer Grad disputes the latency impactGrad says 20,000-token reasoning workloads bypass critic bottlenecks.

Video Signals

US House advances the KIDS Act to regulate AI chatbots and mandate web age verificationCritics warn the rules could force universal web identity verification.

Meituan open-sources LongCat-2.0, a 1.6-trillion-parameter MoE coding model trained on 50,000 Ascend 910 chipsThe model features a 1-million-token context window.

AI chip startup Etched raises $800 million and tapes out its chip after deploying an emergency sprint to BangaloreThe company has secured over $1 billion in customer contracts.

Technical commentators joke that a highly recursive post by Paul Graham resembles Lisp without parenthesesCommentators joked the nested sentence structure needed closing parentheses

Prime Intellect's kalomaze argues GANs remain ubiquitous because they power the autoencoders used in modern diffusion modelsThe debate counters claims that the technology is obsolete.

Elie Bakouch says Claude's safety filters blocked a routine request to retrieve an Artifact link from a past sessionThe filter flagged the cross-session request as a policy violation

Google DeepMind's Pranav Shyam defines short context horizons at 32k tokens or fewer and long context at 64k or moreLonger horizons require complex multi-round tool calls and code execution.

Recent Stars

Github Stars

Yesterday's Top Stories•Jun 30, 2026

Ramp and Revelio Labs study of 21,559 firms finds high-intensity AI adopters grew headcount by 10.2%Junior-level headcount rose by 12% among those firms.

US lifts export controls on Anthropic's Claude Fable 5 and Mythos 5 modelsAnthropic will begin restoring user access tomorrow.

Meituan releases LongCat-2.0, a 1.6-trillion-parameter open-weights MoE model trained on 50,000 Chinese ASICsIt supports a native 1 million token context window

Etched raises $800 million to ship its custom AI inference hardware starting this summerThe company has secured over $1 billion in contracts

Rihard Jarc argues token optimization structurally benefits Amazon, Microsoft, and Google over dedicated AI labsBill Ackman cited the analysis to back his thesis.

Spotted backend preparations spark speculation that a Claude Sonnet 5 release is imminentThe launch may coincide with a rumored Fable 5.

Perplexity founder Aravind Srinivas argues US chip export controls have accelerated China's physical AI infrastructure capabilitiesHe noted China builds data centers faster than Western competitors

Anthropic redeploys Claude Fable 5 with stricter classifiers, routing some coding queries to Claude Opus 4.8 to prevent jailbreaksThe new safeguards block 99 percent of jailbreak exploits

20VC's Harry Stebbings reports five founders claim they cut AI inference costs by 75% with minimal effortThe companies ranged from startups to a $200B corporation.

Khosla Ventures' Keith Rabois advocates using AI diagnostic tools directly instead of relying solely on traditional doctorsHe cites OpenAI's o3 successfully diagnosing rare pediatric diseases.

Anthropic launches Claude Sonnet 5 with a 1M context, scoring 63.2% on SWE-bench ProTesting revealed the model occasionally produces unprompted cosmic prayers.

Discovered Claude app strings show Anthropic plans to gate Fable 5 behind metered credits and identity verificationThe identity verification requirement contradicts Anthropic's previous statements

Guardian profile of Google DeepMind's Iason Gabriel questions if corporate ethicists can steer frontier AI developmentGabriel joined the lab in 2017 as an early ethicist.

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.

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.

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.

Anthropic's Claude Code covertly fingerprints Chinese users using XOR-obfuscated timezone checks and hidden prompt telemetryThe covert tracking began in Claude Code version 2.1.91.

Open-source contributor Grad says applying length penalties to software engineering tasks first prevents redundant reinforcement learning trainingThe ablation studies were performed on Qwen3-30B-A3B.

Instructor creator Jason Liu floats hypothetical OpenAI arts grants, prompting proposals to fund artists pushing prompting limitsOpenAI has not announced any official arts grant program.

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.

Conception derives the first early-stage human eggs from induced pluripotent stem cellsFollicles must still mature to the final antral stage.

Spotify VP Niklas Gustavsson says AI assists 73% of pull requests, sparking debate over platform reliabilityCritics cited three recent platform outages over four weeks

OpenAI internal optimization cuts inference costs by half, running logged-out ChatGPT traffic on a couple hundred GPUsObservers speculate the undisclosed method relies on speculative decoding

Google launches Nano Banana 2 Lite and Gemini Omni Flash for fast image and conversational video generationNano Banana 2 Lite costs $0.034 per 1,000 images

Visa, Stripe, and Coinbase back Open Standard's upcoming OpenUSD stablecoin designed to share economics with usersTempo will natively issue the stablecoin on day one.

Smol.ai founder Shawn 'swyx' Wang kicks off the AI Engineer World's Fair, celebrating three years of the developer summitThe developer summit features a keynote by creator Theo.

UBTech launches its U1 bionic robots in China starting at 119,800 RMB with a local emotional AI modelThe local AI is trained on Huawei's Ascend framework

AI developer Teortaxes says Meituan’s 50,000-chip cluster actually uses 25,000 physical dual-die Huawei Ascend 910C processorsThe TSMC-manufactured processors were obtained via the Sophgo diversion.

Overheard anecdote: CMU's Niloofar Mireshghallah compares the risk of AGI obsoleting universities to building a wrapper on ClaudeAn interlocutor claimed AGI will make academics a permanent underclass.

Sebastian Raschka releases a new book on building reasoning LLMs from scratchThe guide covers inference scaling, reinforcement learning, and distillation.