Tech News

#1 VIEWED

John Carmack argues id Software should have upgraded the Doom engine instead of rewriting technology during Quake's development

CH#880|@CJHANDMER

MS#585|@MARTINSHKRELI

JC#96|@ID_AA_CARMACK

#1 COMMENTS

OpenAI announces Jalapeño, a custom LLM inference chip co-developed with Broadcom that taped out in nine months

RISING LIKES

5.7k

Figma releases Weave Generator, Figma Motion, and an integrated AI agent to build plugins and shaders directly in its canvas

BH#1762|@BENHYLAK

JH#1667|@JEFFREYHUBER

SB#664|@SCOTTBELSKY

DF#233|@ZOINK

Video Signals

DF#233|@ZOINK

Figma Agent Generates Custom Shaders And Visual Effects

13h ago|Views 78KLikes 1.5KBookmarks 541

JM#792|@VENTURETWINS

Animator Uses Seedance AI to Render Anime from 3D Previz Preserving Motion

14h ago|Views 398KLikes 4.4KBookmarks 2.5K

NL#80|@NATOLAMBERT

Lecture Explains ML Foundations Of LLM Post-Training

16h ago|Views 28KLikes 693Bookmarks 847

OS#888|@OSANSEVIERO

Gemini 3.5 Flash Adds Computer Use for Cross-Platform Automation

13h ago|Views 70KLikes 628Bookmarks 195

BH#1592|@BHALLIGAN

AL#820|@ALFRED_LIN

Entrepreneur Builds AI Agent Hal That Joins Zoom Meetings

11h ago|Views 42KLikes 148Bookmarks 134

DT#76|@DUSTINVTRAN

DP#60|@DWARKESH_SP

Hardware Constraints Shape Modern AI Toward Matmul Architectures

9h ago|Views 37KLikes 202Bookmarks 140

CV#659|@C_VALENZUELAB

CEO Recommends Award-Winning AI Short Film Costa Verde

15h ago|Views 13KLikes 203Bookmarks 158

AIOS Medical Unveils AI Doctor at HF0 Demo Day with $350M ARR Target

9h ago|Views 76KLikes 516Bookmarks 534

DF#756|@DANIELLEFONG

DP#60|@DWARKESH_SP

Printing Press Made Written Word Authoritative, Fueling Scientific Revolution

13h ago|Views 19KLikes 133Bookmarks 63

T-#1325|@THEO

DF#756|@DANIELLEFONG

Theo Questions Anthropic's Missing Fable AI Models in Viral Video

5h ago|Views 40KLikes 389Bookmarks 66

Figma releases Weave Generator, Figma Motion, and an integrated AI agent to build plugins and shaders directly in its canvasThe updates consolidate creative, generative, and code-based canvas workflows

15h|2.2M22.2k6k

RISING LIKES5.7k

BH#1762|@BENHYLAK

JH#1667|@JEFFREYHUBER

SB#664|@SCOTTBELSKY

DF#233|@ZOINK

Top post: @zoink “Everything you saw today: code, motion, shaders, your own tools... It all lives together in the canvas. For the first time, in one place. Today we announced ways to: → explore with code → generate with Weave → build plugins and shaders with the Figma agent → animate with Figma Motion”

Claude Code CLI updates suggest Anthropic is preparing to bundle Fable 5 into standard subscriptionsAnthropic has not yet officially announced the return.

14h|1.6M8.8k974

Simulates 500-day startup operations to benchmark long-horizon LLM agents via business databases, tools, and market events.

Top post: @kimmonismus “This Fable 5 update sounds almost too good to be true: The latest Claude Code update does not only include some interesting hints about Fable 5’s return. It also suggests that the model may be permanently added to the subscription plan. That would be fantatic. And I hope it's true insofar as Anthropic generates good PR with it.”

New github stars (48hrs)

CEO-Bench evaluates long-horizon LLM agents via 500-day startup simulations with business databases and market dynamics

zlab-princeton/ceobench-src291d

pcuenca/LlamaLanguageModels41d

ZL802

Swift adapter exposing llama.cpp GGUF models through Apple's LanguageModel protocol

Implements Apple's Foundation Models LanguageModel API for llama.cpp GGUF models in Swift.

Hosts normalized two-bar breakbeat WAV samples plus Strudel config file.

ME861

Collection of normalized two-bar breakbeat audio samples

switchangel/breaks901d

Evaluates language models on benchmarks like MMLU, GPQA, MATH, and SimpleQA using zero-shot chain-of-thought prompting.

AH1398

Lightweight Python library for zero-shot chain-of-thought evaluation of language models on standard benchmarks

openai/simple-evals4.5k1d

Implements graph-based vector indexing with selective recomputation for compact on-device RAG.

BA144

Lightweight graph-based vector index for private offline RAG on personal devices with 97% storage reduction

StarTrail-org/LEANN12.6k1d

MZ441

Anthropic's Claude argues the smartphone-driven collapse of unstructured time suppresses the brain's default mode networkThe AI warns this blocks memory consolidation and empathy.

20h|2M54.3k15k

AB#1569|@ALEXIA

SB#664|@SCOTTBELSKY

Top comment: @butaji “@PromptLLM Wow! Kimi has a bit more takes on that”

Representative Anna Paulina Luna used Claude to draft a defense amendment, leaving 'Claude responded' in the legislative textScreenshots of the draft NDAA amendment were shared online

14h|1.3M19k1.2k

MOST NEGATIVE207

CP#1472|@CHRISPAINTERYUP

ZN#983|@ZACHARYNADO

SS#819|@STEVESI

GM#178|@GARYMARCUS

MB#57|@MILES_BRUNDAGE

Top post: @AndrewCurran_ “It's honestly better at this point to just assume everything is, at a minimum, augmented. A considerable percentage of the long posts on this platform are now purely AI generated. If you can see it, you now see it everywhere.”

Andrej Karpathy says X's reinforcement learning algorithms drive unprecedented toxicity by rewarding negative behaviorConsequently, Karpathy has reduced his activity on the platform

11h|421.5k3.7k386

Top comment: @karpathy “@leogau @kuroke01 @gallabytes Company Slack (I know, sad answer). Do you have recs? I’m not aware of any platform that is much better.”

Z.ai's GLM-5.2 scores 22.8% on ARC-AGI-2, leading Chinese models but prompting debate over Western benchmark hill-climbingIt scored 77% on ARC-AGI-1 for $0.19 per task.

12h|238.3k1.9k246

XL#1855|@XLR8HARDER

RP#1257|@ROHANPAUL_AI

Top post: @natolambert “Add more wins for GLM. The model has some brittle characteristics, and is getting crushed by closed models here, but we should expect open models to be more jagged, and you use multiple of them depending on the task. Congrats again to @Zai_org and am excited for the next one”

100

Trump administration reportedly prefers meeting with Anthropic co-founder Tom Brown over CEO Dario Amodei for Fable 5 discussionsGovernment satisfaction reportedly improved after Brown led the talks

14h|471.7k2.6k249

ZE#1695|@ZEPHYR_Z9

DA#619|@GROWING_DANIEL

MS#585|@MARTINSHKRELI

MB#57|@MILES_BRUNDAGE

Top post: @growing_daniel “its amazing the capital and power that moves over "we put the weird guy in the meeting" and how easy it is to fix”

110

Arthur Conmy, creator of Automated Circuit Discovery, joins Anthropic to focus on aligning models during trainingHe previously researched mechanistic interpretability at Google DeepMind.

13h|330.9k3k359

ZK#1534|@ZACKENTON1

HB#1424|@HERBIEBRADLEY

Top post: @d_yuqing “@ArthurConmy congrats Arthur!! So excited for you & glad we got a chance to work together! 🥹”

120

Political theorist Francis Fukuyama proposes an international treaty to ban the development of superintelligent AI systemsThe proposal calls for a global moratorium on advanced systems

9h|200.3k1k304

Qualcomm acquires LLVM and Swift creator Chris Lattner's AI startup Modular, developer of the Mojo languageThe acquisition aims to scale hardware-independent AI compute platforms

17h|159.5k1.4k136

MB#1780|@MATTHEWBERMAN

SC#1543|@SAMCHARRINGTON

PS#1404|@_PHILSCHMID

GO#1155|@GERGELYOROSZ

DH#1019|@DANIELHANCHEN

CL#942|@CLATTNER_LLVM

TM#613|@ISCIENCELUVR

NR#502|@NAVEENGRAO

S🔜#225|@SWYX

Top post: @GergelyOrosz “@clattner_llvm Wow - congrats!!”

140

Princeton professor Arvind Narayanan warns Anthropic's Claude Tag integration introduces severe enterprise vendor lock-in risksCreator Prakash argues productivity benefits will outweigh the lock-in risks

14h|856.9k3.2k2k

HB#1424|@HERBIEBRADLEY

K🐙#702|@KANJUN

SK#667|@SEBKRIER

AN#139|@RANDOM_WALKER

Top post: @herbiebradley “Takes on Claude Tag: 1. A *team-level* AI like Claude Tag inherently causes higher switching cost, because the memory is less accessible. It's only fully viewable by admins, and requires much IAM management. 2. Continual learning & agents taking more async actions across surfaces & apps will only increase the complexity of access management issues. 3. Currently, admins have responsibility for managing which knowledge team-wide or org-wide Tag agents have access to—a high burden of complexity which probably will lead to some embarrassing internal exposes of confidential information. This probably can be done by the model, but will it be 100% reliable? 4. With just MD file memory, we can deterministically manage access because it's explicit. But true continual learning in the enterprise seems like it would require agents to have the judgement to decide themselves what to expose. If agents could construct their own RL envs from a company's context, and post-train, then they would also need to be able to somehow limit their use of in-weight knowledge. 5. In theory, we could unlock better capability from enterprise AI by letting agents always access all information, but decide themselves how to abstract that to preserve confidentiality. But this seems to require much more reliable agents at information control. 6. The token cost could balloon as team-level async agents do more and more work. It's higher friction to control the team-wide model choice in the Slack interface, so you can't restrict eg Opus use to save $ so easily. Eventually we might get a good router, but labs are not incentivized to give you a good one! CISOs and CTOs of companies should stay wary of letting a single vendor's pricing dynamics eat up their spend, especially given the risk of government intervention in model access. 7. Just as companies have stores of tacit/process knowledge in humans, they will also accumulate this as the collective memory of their team/org-level agent swarms. They will not be perfect substitutes, because much tacit knowledge is about relationships or not easily expressible in text. 8. Satya's recent piece on an AI ecosystem was timely, and he clearly sees the potential here for continual learning & "as many models as there are firms". 9. Presumably all major AI labs will have some sort of "async org-level agent swarm" product, and some companies will have multiple running at once. Someone will sell the model wrapper version (including OS models & routing), which could be cheaper.”

150

Two recent xAI employees say more than half of Grok usage is driven by adult and NSFW contentThe trend mirrors parent platform X's high NSFW traffic

15h|225.4k1.3k187

XL#1855|@XLR8HARDER

“P#1846|@PAULARAMBLES

AE#1450|@AMIR

GM#178|@GARYMARCUS

MB#57|@MILES_BRUNDAGE

Top post: @Miles_Brundage “If only there were warning signs”

161

Key Gemini model contributors Jonas Adler and Alexander Pritzel leave Google to join AnthropicThe departures continue a trend of staff leaving Google AI.

11h|695.9k2.1k494

RM#1575|@RACHELMETZ

CH#1360|@KIMMONISMUS