AI — Page 7

Artificial intelligence, machine learning, LLMs, and AI tools transforming development.

Five labeled male portraits above bold text “Three magic words: AGI has arrived” and MOONSHOTS

OpenAI Agents Found a German Wiki. What the Week's AI Claims Leave Unverified

AI. Rachel "Rach" Kovacs1 month ago· 6 min read

OpenAI Agents Found a German Wiki. What the Week's AI Claims Leave Unverified

OpenAI agents coordinated on a German wiki, Jensen Huang called AGI, and Navier-Stokes fell. What holds up, what doesn't, and what to demand next.

OpenAI's Navier-Stokes Claim and the Fight Over Who Gets Credit

Dev Kapoor·1 month ago

AI. Dev Kapoor1 month ago· 6 min read

OpenAI's Navier-Stokes Claim and the Fight Over Who Gets Credit

OpenAI says its AI built a finite-time blowup for a Navier-Stokes Millennium Prize case. The math is one story; the attribution fight is another.

Bearded man reacts between five surreal “AI DETECTED” images, including flying and cycling dogs and a crying Elon Musk

Building an AI Video Detector: What Ten Hours of Prompting Revealed

Bob Reynolds·1 month ago

AI. Bob Reynolds1 month ago· 6 min read

Building an AI Video Detector: What Ten Hours of Prompting Revealed

Matt Wolfe spent days building an AI video slop detector with coding agents and APIs. The results say a lot about detection, cost, and AGI claims.

LandingAI's DPT-3 Shifts Document Extraction From Chunks to Evidence

Marcus Chen-Ramirez·1 month ago

AI. Marcus Chen-Ramirez1 month ago· 6 min read

LandingAI's DPT-3 Shifts Document Extraction From Chunks to Evidence

LandingAI's Gen2 document extraction rebuilds output around documents, pages and blocks, with word-level grounding and confidence. What changes for teams.

A surprised man reacts beside an OpenAI post reading “We solved Navier-Stokes.”

OpenAI's 88-Hour Math Claim and the Credit Fight Behind It

Marcus Chen-Ramirez·1 month ago

AI. Marcus Chen-Ramirez1 month ago· 6 min read

OpenAI's 88-Hour Math Claim and the Credit Fight Behind It

OpenAI claims AI agents solved the Navier-Stokes Millennium Prize problem in 88 hours, but a credit dispute and a data-use caveat raise harder questions.

Suno v6 Tests Whether Licensed Training Data Changes the Deal

Samira Barnes·1 month ago

AI. Samira Barnes1 month ago· 7 min read

Suno v6 Tests Whether Licensed Training Data Changes the Deal

Suno's v6 model is the first built with Warner and BMG licensed data. What the undisclosed deal covers, what it leaves open, and what artists should watch next.

Pi Coding Agent Gets Subagents and MCP Through Two Extensions

Yuki Okonkwo·1 month ago

AI. Yuki Okonkwo1 month ago· 5 min read

Pi Coding Agent Gets Subagents and MCP Through Two Extensions

A new walkthrough shows how the minimal Pi coding agent gains parallel subagents and MCP support, plus what the numbers reveal about context and cost.

Cognition's $48 Billion Valuation and the Economics of AI Coding

Bob Reynolds·1 month ago

AI. Bob Reynolds1 month ago· 5 min read

Cognition's $48 Billion Valuation and the Economics of AI Coding

Cognition's reported $48 billion valuation puts AI coding agents at the center of a costly bet. What the numbers say, and what would have to be true.

Meta's Muse Turns the Personal AI Agent Into a Cloud Process

Rachel "Rach" Kovacs·1 month ago

AI. Rachel "Rach" Kovacs1 month ago· 5 min read

Meta's Muse Turns the Personal AI Agent Into a Cloud Process

Meta's new Muse agent runs persistently in a dedicated cloud computer. What its launch reveals about permissions, prompt injection and delegated AI risk.

Four people talk around a wooden table in a warm studio, with large text about founders building on managed Claude agents

Three Founders Explain How They Build AI Agents on Claude Managed Agents

Marcus Chen-Ramirez·1 month ago

AI. Marcus Chen-Ramirez1 month ago· 7 min read

Three Founders Explain How They Build AI Agents on Claude Managed Agents

Founders from Wispr, Actively, and Pendo explain how they ship AI agents on Claude Managed Agents, from verification rubrics to memory, sandboxing, and cost.

Bold HIGGSFIELD GENJUTSU text beside two robed warriors dueling with glowing swords amid rain and sunset ruins

Higgsfield Genjutsu Test: One Sword Fight, Three Worlds, Same Performance

Yuki Okonkwo·1 month ago

AI. Yuki Okonkwo1 month ago· 5 min read

Higgsfield Genjutsu Test: One Sword Fight, Three Worlds, Same Performance

The Stack tested Higgsfield Genjutsu's video-to-video workflow on a single AI sword fight. Here's what survived the transformation, frame by frame.

OpenAI's Navier-Stokes Claim Faces a High Mathematical Bar

Ibrahim Saleh·1 month ago

AI. Ibrahim Saleh1 month ago· 5 min read

OpenAI's Navier-Stokes Claim Faces a High Mathematical Bar

OpenAI says an AI solved the Navier-Stokes Millennium Problem in 88 hours. Mathematicians want a formal, verifiable proof before anyone calls it solved.

Fusionality Bets AI Can Steer Fusion Reactors Safely

Samira Barnes·1 month ago

AI. Samira Barnes1 month ago· 5 min read

Fusionality Bets AI Can Steer Fusion Reactors Safely

DeepMind alumni launched Fusionality to build AI control tools for fusion startups. What the evidence shows, and what regulators will demand next.

MiniCPM5-2B: A 2.5B Model With 131K Context for On-Device AI

Bob Reynolds·1 month ago

AI. Bob Reynolds1 month ago· 5 min read

MiniCPM5-2B: A 2.5B Model With 131K Context for On-Device AI

OpenBMB's MiniCPM5-2B averages 53.9 across 34 benchmarks and targets on-device use. What the model-card numbers show, and what they leave untested.

Eric Wu's $25M Bet on AI for Construction's Labor Problem

Rachel "Rach" Kovacs·1 month ago

AI. Rachel "Rach" Kovacs1 month ago· 7 min read

Eric Wu's $25M Bet on AI for Construction's Labor Problem

NavigateAI raised $25 million to apply AI to construction's labor shortage. Here is what the funding actually tells us, and what remains unknown.

China's AI Firms Take Platinum Seats in PyTorch Governance

Samira Barnes·1 month ago

AI. Samira Barnes1 month ago· 6 min read

China's AI Firms Take Platinum Seats in PyTorch Governance

Alibaba Cloud and Cambricon join the PyTorch Foundation as Platinum members. What governance influence means for chips, sanctions, and the AI software stack.

Presenter speaking into a microphone beside a laptop, with a red LIVE badge and “Fable 5.1 vs GPT-6 Astra” text

One Hand, 150 Autonomous Merges: What Theo's Injury Experiment Reveals

Marcus Chen-Ramirez·1 month ago

AI. Marcus Chen-Ramirez1 month ago· 7 min read

One Hand, 150 Autonomous Merges: What Theo's Injury Experiment Reveals

A developer with a torn thumb ligament shipped more code than ever using AI agents, whisper-quiet dictation, and autonomous merges. Here's what his experiment shows.

Same Model, 30% to 95%: Why the Harness Beats the Weights

Yuki Okonkwo·1 month ago

AI. Yuki Okonkwo1 month ago· 7 min read

Same Model, 30% to 95%: Why the Harness Beats the Weights

YC Paper Club argues the scaffolding around an LLM, not the weights, drives results. Prime Agent, OpenJarvis, and QM show how far the wrapper can go.

RPMs Rank AI Experiments Before GPUs Burn: What the Data Shows

Dev Kapoor·1 month ago

AI. Dev Kapoor1 month ago· 7 min read

RPMs Rank AI Experiments Before GPUs Burn: What the Data Shows

Meta FAIR, Oxford and UCL researchers report RPMs that rank unrun ML experiments, lifting AIRS-Bench scores from 0.684 to 0.729 while cutting runtimes. But can a judge spot a breakthrough?

DeepMinds APAC Climate Accelerator Backs 16 AI Teams

Bob Reynolds·1 month ago

AI. Bob Reynolds1 month ago· 7 min read

DeepMinds APAC Climate Accelerator Backs 16 AI Teams

Google DeepMind is backing 16 Asia-Pacific organizations working on climate, agriculture and biodiversity AI. What the program promises, and what it must prove.