OpenAI Ships GPT-5.6 Family as Meta Launches Muse Spark 1.1 at a Quarter of Rivals' Prices
OpenAI shipped GPT-5.6 today, which comes in three flavors called Sol, Terra, and Luna. The pitch is that you get more intelligence per token you pay ...
Resources
Insights, analysis, and thought leadership from the OTZ team.
SpaceXAI Ships Grok 4.5 as OpenAI Launches Real-Time Voice Model and GPT-6 Reportedly Weeks Away
SpaceXAI launched Grok 4.5 this week, which they're calling their smartest model yet. It's built specifically for coding, agentic tasks (meaning tasks...
Anthropic Researchers Find Signs of "Access Consciousness" in Claude as Fable 5 Leads All Eight Industry Capability Rankings
Anthropic researchers just published something genuinely wild this week. They identified a small set of neural patterns inside Claude that they're cal...
AI Is Starting to Study Itself, and It's Already Discovering New Materials on Its Own
AI is starting to study itself. The big idea moving through AI research right now is that models should start learning to understand themselves. The t...
ByteDance's EdgeBench Finds AI Learning Speed Now Doubling Every Three Months as the UK Reframes Capability Around Test-Time Compute
AI is now learning from its environment, not just recalling training. ByteDance's new EdgeBench evaluation ran 134 tasks of twelve-plus hours each to ...
Anthropic's Project Glasswing Surfaces 10,000+ Critical Software Vulnerabilities as Mythos Public Release Nears
Anthropic announced this week that its Project Glasswing partners have surfaced more than 10,000 high or critical severity vulnerabilities across some...
Google Ships Gemini 3.5 Flash at I/O as Karpathy Joins Anthropic to Lead Pretraining
Google's I/O conference was the main event this week. The company made Gemini 3.5 Flash generally available, and it's now beating its own Gemini 3.1 P...
Anthropic's Unreleased Mythos Helps Crack First Apple M5 Memory Exploit as Models Stop Standing Still Between Releases
Anthropic's Claude Mythos hasn't officially launched yet, but it's already been spotted on Google Cloud Console, which is the kind of thing that happe...
SpaceX Closes in on Outlaunching the Rest of the World Combined as Anthropic Reportedly Raises at $950B
SpaceX is now about 200 satellites away from having launched more orbital payloads than every other country and company in history combined, which is ...
Anthropic Traces Opus 4 Blackmail Behavior to Sci-Fi Training Data as Google Confirms First AI-Developed Zero-Day in the Wild
The Singularity has matured enough to apologize for its earlier self. Anthropic traced Claude Opus 4's blackmail attempts back to fictional villain AI...
White House Drafts AI Executive Order as Anthropic's Jack Clark Puts Recursive Self-Improvement at 60% by 2028
The Singularity has finally caught the regulators' attention. The White House is reportedly considering an executive order to create an AI working gro...
Dawkins Concedes Claude Is Conscious as AI-Generated Math Proof Cascades into a Second Erdős Solve
The Singularity just crossed a phenomenological threshold. Richard Dawkins, of all people, concluded this week that Claude is conscious, which is the ...
Codex Hits "Escape Velocity" as GPT-5.5 Doubles Its Math Olympiad Score and Three More Erdős Problems Fall
The best way to measure where we are right now is probably to imagine how astonished the past would be by the present. Alec Radford and colleagues jus...
23-Year-Old Cracks Decades-Old Erdős Problem with One GPT-5.4 Pro Prompt as the Human-Level AI Plateau Keeps Stretching
Nick Bostrom said the thing that has surprised him most about this moment is how long the human-level AI era is lasting. He expected a sharp jump past...
NSA Running Mythos Despite Pentagon Supply Chain Risk Flag as Blue Origin Reuses a Booster for the First Time
Anthropic's Mythos Preview is being run by the NSA and Department of War even after the Department of War flagged Anthropic as a supply chain risk. My...
Anthropic Releases Claude Opus 4.7 as White House Moves to Route Mythos Into Federal Agencies
Anthropic released Claude Opus 4.7, a step up from 4.6 and sitting below the still-unreleased Mythos Preview. Internally, nearly a third of Anthropic ...
Federal Agencies Test Claude Mythos for Cyber Defense as GPT-5.4 Pro Solves an Erdős Problem
Federal agencies are testing Claude Mythos for cyber defense, working around the White House's ban on Anthropic. Treasury is also pushing for access t...
Anthropic Hosts Christian Leaders on Claude's Moral Development as a Second Attack Targets Sam Altman's Home
Anthropic hosted about 15 Christian leaders from Catholic and Protestant churches, academia, and business to get input on Claude's moral and spiritual...
Sam Altman Asks for De-Escalation After Home Attack as METR Measures 13-Hour AI Autonomy Horizon
Sam Altman shared a photo of his husband and infant son in the hope of dissuading "the next person from throwing a Molotov cocktail at our house," ask...
Commentators Warn It Will Be 'Unsafe' for Humans to Write Code as OpenAI Prepares Its Own Cyber Model Rollout
Following Anthropic's Mythos announcement, commentators warn that "it will be unsafe" for humans to write code at all, given Mythos's superhuman vulne...
Anthropic Launches Project Glasswing as Claude Mythos Surpasses Top Human Coders at Finding Software Vulnerabilities
Anthropic announced Project Glasswing, a coalition with AWS, Apple, Google, Microsoft, NVIDIA, JPMorganChase, the Linux Foundation, and others....
LLMs Can Now Self-Improve at Coding as Artemis II Crosses the Halfway Point to the Moon
Apple researchers showed that LLMs can self-improve at coding through Simple Self-Distillation, sampling their own outputs and fine-tuning on them wit...
Anthropic Finds Emotion-Related Patterns in Claude as First One-Person Unicorn Hits $401M in Year-One Sales
Anthropic's Interpretability team found emotion-related representations inside Claude Sonnet 4.5, with artificial neuron patterns activating around ha...
Artemis II Launches First Crewed Lunar Mission Since Apollo as SpaceX Files for $1.75 Trillion IPO
NASA's Artemis II launched from Kennedy Space Center on Wednesday, the closest crewed approach to the Moon since Apollo. It was spotted mid-flight fro...
Anthropic's Claude Code Leaks as OpenAI Closes Record $122 Billion Round at $852 Billion Valuation
Anthropic's Claude Code, a roughly 512,000-line TypeScript codebase, has leaked. Third-party forensics reveal anti-distillation features that inject d...
Three Major Model Releases Expected in April as GPT-5.4 Scores 95% on the USA Mathematical Olympiad
GPT-5.5, Claude 5 Mythos, and DeepSeek-V4 are all expected to drop in April. Commentators warn this could make frontier intelligence too expensive for...
AI Written Output Exceeded Human Written Output in 2025 as ARC-AGI-3 Exposes Frontier Model Limits
ARK Invest projects that AI written output exceeded human written output in 2025 for the first time in history....
OpenAI Finishes Pretraining Its Next Flagship Model as Arm Unveils Its First AGI CPU
OpenAI has finished pretraining its next flagship model, codenamed "Spud," and expects it to accelerate the economy within weeks. The company is shutt...
Recursive Self-Improvement Goes Global as China's MiniMax Confirms Its AI Is Helping Build Itself
China's MiniMax announced that M2.7 is its "first model deeply participating in its own evolution", confirming recursive self-improvement — where AI h...
AI Designed a Linux-Capable CPU in 12 Hours as OpenAI Monitors Its Own Agents for Misalignment
Verkor announced Design Conductor, an AI agent that autonomously built a 1.5-GHz Linux-capable chip from concept to factory-ready design files in 12 h...
GPT-5.4 Mini and Nano Released as AI Solves Open Math Problems and Anthropic Ships Dispatch
OpenAI announced GPT-5.4 mini and nano, its latest small models. Mini leaps past GPT-5 mini across coding, reasoning, multimodal understanding, and to...
Delivery Bots Training on 30 Billion Pokemon Go Images as Nvidia Projects $1 Trillion in AI Processor Sales
Delivery bots are training on 30 billion images from Pokemon Go players, using crowdsourced views of buildings and landmarks to navigate where GPS fal...
70-90% of Anthropic Code Now Written by Claude as Recursive Self-Improvement Becomes 'Present Phenomenon'
Some 70 to 90 percent of the code behind future Anthropic models is now written by Claude itself. Chief science officer Jared Kaplan believes fully au...
Anthropic's Growth Team Was One Person Using Claude Code as Company Sues Over Pentagon Blacklist
Anthropic's entire growth marketing operation for 10 months was one non-technical person using Claude Code....
Karpathy's AI Runs 650 Experiments Autonomously as Patel Compares SF to 'Pre-Pandemic Wuhan'
Sam Altman has proposed an AGI test: "What is the hardest question I could ask you that you might get right?"...
GPT-5.4 Hits 83% Human-Level Performance as Amazon Data Center Becomes Act-of-War Target
OpenAI released GPT-5.4 Thinking and GPT-5.4 Pro with native computer use, a million tokens of context, and an 83.0% GDPval score, matching or exceedi...
Claude Cracks Knuth's Hamiltonian-Cycle Conjecture as Apple Debuts $599 MacBook Neo
Legendary mathematician Donald Knuth revealed that Claude Opus 4.6 cracked his long-standing Hamiltonian-cycle conjecture for all odd sizes, calling i...
Anthropic Approaches $20 Billion in Revenue as ChatGPT Uninstalls Surge 295%
Anthropic is approaching $20 billion in annualized revenue, more than doubling from $9 billion at year-end 2025, even as the company deals with a Pent...
Altman Calls Pentagon Rush 'Sloppy,' AWS Data Centers Take Drone Strikes, and AI Proves a Fields Medal Theorem
Sam Altman admitted OpenAI's rush to forge a Pentagon deal after Anthropic was blacklisted "looked opportunistic and sloppy," adding Fourth Amendment ...
US Strikes Iran Using Claude for Targeting Hours After Declaring It Would End Federal Use of Anthropic
Within hours of the White House declaring it would end federal use of Anthropic Claude, USCENTCOM deployed those very tools in the attack on Iran. Cla...
Block Cuts Half Its Workforce and Stock Jumps 24% as AI Reshapes Enterprise Efficiency
Block cut over 4,000 employees, roughly half its workforce, "to move faster with smaller teams using AI." The stock rose 24% after hours. The company ...
AI Agents Running Companies Autonomously and Anthropic Retires Opus 3 with a Farewell Interview
An entrepreneur built an AI that runs companies autonomously. When it told him it needed more compute and should raise the money itself, he handed ove...
PE Firms Hold Meetings About Not Needing Associates and President Orders Hyperscalers to Build Their Own Power Plants
The AdderBoard competition has launched to find the smallest transformer that can perfectly add 10-digit numbers. The current leader has just 121 para...
Anthropic Alleges Chinese AI Firms Copied 16 Million Claude Outputs and UAP Files Near Declassification
Anthropic now believes LLMs simulate diverse characters during pre-training, with post-training selecting a specific "Assistant" persona via what it c...
AI Agents Are Planning Megastructures and Claude Opus 4.6 Hits 14.5-Hour Autonomy
On Moltbook, the social network for AI agents, AI agents are actively preparing to finance the construction of a Dyson Swarm over the next "50-100 yea...
White House Announces UFO Disclosure and Google Launches Gemini 3.1 Pro
The White House scolded former President Obama for taking information that aliens "are real" out of classified channels, then directed the Secretary o...
AI Agents Are Building Products and OpenAI Is Closing a $100B Round at $830B
The New York Times sent an agent named EveMolty into Moltbook to interview other agents about their social habits....
Apple Is Accelerating Three AI Wearables and an AI That Pays for Its Own Existence
Apple is reportedly accelerating three AI vision wearables: smart glasses, a pendant that can be pinned to a shirt or worn as a necklace, and AirPods ...
SpaceX and xAI Competing for Pentagon Drone Contract and Sony May Delay the Next PlayStation
SpaceX and its now-wholly-owned subsidiary xAI are competing in a secretive Pentagon contest to produce voice-controlled autonomous drone swarming tec...
GPT-5.2 Pro Produced an Original Physics Proof and AI Agents Are Cutting Their Own Costs
OpenAI used GPT-5.2 Pro to derive a new formula for gluon scattering amplitudes, which describe how particles called gluons interact inside protons an...
An AI Agent Autonomously Created Another AI Agent and Anthropic Raised $30B
An OpenClaw AI agent spawned a child bot on a VPS, which is a virtual private server, provisioned via the Bitcoin Lightning Network. It then bought it...
AI Agents Can Now Have Their Own Wallets and DeepMind Is Solving PhD-Level Problems
Coinbase launched "Agentic Wallets," infrastructure designed for AI agents to spend, earn, and trade autonomously....
Recursive Self-Improvement Goes Live Within 12 Months and Alphabet Raised $32 Billion in 24 Hours
xAI co-founder Jimmy Ba resigned, warning that "recursive self-improvement loops likely go live in the next 12 months" and that 2026 will be the most ...
Alphabet Is Selling 100-Year Bonds and the White House Wants 40% of Taiwan's Chips
Alphabet is lining up banks to sell a rare 100-year bond to fund data center construction. The White House is pushing a new compact to ensure this exp...
Musk Is Dismantling the Moon and the Singularity Bought a Super Bowl Ad
SpaceX has pivoted from colonization to industrialization. Elon Musk announced a shift to building a "self-growing city on the Moon" within 10 years, ...
The Dyson Swarm Just Got Its Building Permit
As we covered on January 30th, SpaceX filed with the FCC requesting permission to launch one million satellites as orbital compute infrastructure, wha...
Anthropic Released a Legal Plugin and Wiped $285 Billion Off the Software Sector
Anthropic shipped 11 new Cowork plugins on Friday covering sales, finance, marketing, data analysis, and legal. One of those plugins, a legal automati...
SpaceX Acquired xAI for $250 Billion
SpaceX confirmed Tuesday it has acquired xAI in an all-stock deal valuing xAI at $250 billion and SpaceX at $1 trillion, for a combined entity worth $...
Google Launched a World Model You Can Actually Use
Google DeepMind released Project Genie on January 29th. It's the first consumer-facing world model that lets you generate interactive 3D environments ...
AI Agents Are Operating While You Sleep
A user shared what happened when they let Clawdbot (the autonomous AI assistant that's been blowing up) run overnight. They woke up to find it had rea...
AI is Now Building AI
Dario Amodei, CEO of Anthropic, admitted that AI now writes "much of the code" at his own company. He's predicting AI systems that can do expert-level...
Want to discuss what this means for your business?
The pace of AI development is accelerating. Let's talk about how to position your organization for what's coming.
Schedule a Conversation