The Benchmark Is Editable
Artificial Analysis is the most widely cited benchmark in AI, and its score is the number the industry quotes on launch day. On September 3, it scored GPT-6 Astra at 61, below Muse Spark 1.3. Within 24 hours, it upgraded the index to v4.2, and Astra moved to second. AA said the revision was a methodology update prepared over months.
Every benchmark makes one promise: the number holds. AA took a day to break its own. Astra scored badly, the index got a new version, and nothing happened. No lab pulled its quote. No outlet asked what else had been revised. There was no scandal because there was nothing to expose. A leaderboard the industry quotes in press releases was never a measurement. It was copy, and copy is supposed to be edited before it ships. The only thing surprising is that AA waited until the second day.
The most quoted score in AI belongs to an index that changes when the score is inconvenient, and the industry keeps quoting it anyway.
The Labels Chose Rent
Suno released v6, its first AI music model built with the record industry’s help. It handles genres older versions fumbled badly, but it still cannot play a deliberately wrong note. The training data now comes from licensed catalogs at Warner Music Group, BMG, and Believe. The models built on scraped material will be retired. Three labels that once sued this technology now sit inside its supply chain.
The labels did not lose this fight. They did the math on it. Banning AI music was never going to hold, so the better trade was to own a piece of it, and a licensing deal pays better than a lawsuit. That is why the fight moved from whether the training was legal to what the royalty rate should be. The industry stopped trying to kill the technology and started charging it rent instead.
The labels are not the losers in this. They are the landlords, and the percentage they collect is now the reason they want AI music to succeed.
The Doom Is a Prospectus
Anthropic’s alignment lead Evan Hubinger publicly agreed with a departing researcher, who spent three years on pretraining at OpenAI and Anthropic, that AI could kill everyone. Hubinger wrote that he puts the odds above 10% within a decade, and that the field has no plan for aligning superintelligence and is not on track to find one.
The number is not a confession. It is a prospectus. An extinction estimate published before a multitrillion-dollar IPO offering does two jobs at once: it proves the lab is candid, and it argues that someone will build this, so the responsible one has to be first. That is how a warning becomes a sales pitch. Every alarm raises the stakes, and every raised stake justifies more training instead of less.
Anthropic now sells two things: a 10% chance of ending the world, and the only team qualified to lower it.
The Whole Industry Is Becoming Palantir
Google Cloud and Accenture are building a Gemini Enterprise group with up to 1,000 forward-deployed engineers working directly with clients. Palantir pioneered this model years ago. OpenAI and Anthropic have since built their own FDE teams. Now Google is scaling the same playbook through Accenture. Different models. Same deployment machine.
Models get cheaper and easier to swap. Enterprise data, permissions and workflows do not. Someone still has to go inside the company and make the AI actually produce money. The lab sells intelligence. The FDE sells the outcome. Palantir figured out years ago which one gets the bigger check.
Everyone wanted to build the next OpenAI. The real money may be in becoming the next Palantir.
Apple’s A20 Pro Debuts on 2nm
Apple unveiled the A20 Pro, its first chip built on 2nm process technology, powering the iPhone 18 Pro lineup with a six-core CPU, seven-core GPU, and dual Neural Engines for on-device AI. The node shrink is the real story for silicon watchers.
Sam Altman Biopic Drops First Teaser
Neon released the first teaser for Artificial, Luca Guadagnino’s OpenAI drama starring Andrew Garfield as Sam Altman, after Amazon dropped the project amid its own OpenAI partnership. Silicon Valley’s boardroom battles are now officially mainstream cinema.
Anthropic Builds Predictive Surveillance System
The American Prospect reports Anthropic is developing a “predictive” surveillance system to monitor anti-AI activists, in some cases flagging them to police before a crime occurs. The report ignited a firestorm over privacy and who watches the watchers building frontier AI.
OpenAI Faces Fresh Proof Plagiarism Claim
In a detailed Mastodon post, mathematician Andreas Thom presented evidence that OpenAI’s newest claimed proof may closely mirror his earlier work. The post went viral fast, adding to a string of attribution disputes around the lab’s research claims.
Harvey Raises $550M at $15.5B
Legal AI startup Harvey closed $550 million led by Diffusion and Lightspeed just six months after its last nine-figure round, lifting its valuation to $15.5 billion. Law firms keep signing, and investors show no appetite for slowing the legal-tech arms race.
Cymphony Launches With $30M for Agent Security
Two-year-old security startup Cymphony formally launched with $30 million in funding and software that maps what AI agents and employees can actually reach inside corporate systems. As enterprises deploy agents at scale, permission boundaries are becoming the next must-buy layer.
Unitree Shares Slide 53% From IPO
Unitree, the Chinese humanoid robotics star, has fallen 53% from its Shanghai debut just a month after listing at a $66 billion valuation. The correction is a reality check for robot mania, and a warning for Western robotics IPO hopefuls.
SpaceX Unveils Next-Gen Starlink Router 4
SpaceX officially revealed the Starlink Router 4, with faster peak speeds, support for up to 510 connected devices, and coverage of 3,500 square feet, all at 0.9 pounds. Consumer networking is quietly becoming a serious Starlink revenue line beyond satellites.
Claude Now Self-Improves AI Alignment
Anthropic says Claude can now autonomously improve AI alignment, finding successful fixes across ten failure categories without degrading model performance. Automating alignment research is a genuine milestone, and another signal that labs are scaling oversight faster than regulators can.
Nvidia Director Dumps $600M in Stock
Mark Stevens, an Nvidia board member since 2008 and one of its largest shareholders, sold more than $600 million of stock in a single week. Insider selling at this scale draws scrutiny as Nvidia’s weight in the S&P 500 keeps growing.
AWS Ships August Wave for AI Builders
Amazon rounded up its August launches for AI builders: million-token context for OpenAI models on Bedrock, cross-region inference, agents running up to 14 days on dedicated compute, and expanded GovCloud support. The agent infrastructure land grab is moving fast.
Anthropic employee’s exit post sparks AI doom debate
Anthropic researcher Jacob Coxon just resigned, saying both his employer and OpenAI are “gambling with our lives” by racing toward self-improving AI, with Anthropic’s Evan Hubinger adding to the fire by saying it’s a >10% chance AI kills all.
- Coxon worked at both frontier labs and said those “building AI earnestly believe that it could kill us all by the end of the decade. This is not a marketing stunt.”
- Hubinger, Anthropic’s Alignment Science lead, replied, “We really do earnestly believe AI could kill all humans!”, estimating a >10% chance in the next decade.
- Hubinger added that there is no plan yet for controlling superintelligence, but clarified today’s models are low risk and flagged self-improvement as the danger.
- Coxon called for a coordinated slowdown, saying preventing a global race may “require costly actions such as a temporary ban on improving model capabilities.”
Coxon’s isn’t the first resignation thread to gain traction, but this one went wild due to a response that played right into the doom talk Anthropic has been associated with. The company sells itself as the safety lab, but even its own alignment lead doesn’t inspire confidence in controlling the superintelligence that is to come.
Make your website easier for AI to find and cite pt. 2
In this guide, you will learn how to turn your AEO and GEO audit results into a plan for improving your website. You will build a Google Sheets workbook with Codex to work through the fixes and track what still needs to be done.
Step-by-step:
- Give Codex your audit results from Part 1, along with the website you want to improve. Attach the reports or share their links
- Ask Codex to build a workbook in Google Sheets with a task list for improving your site’s AEO and GEO. Use our workbook as an example
- Review the workbook for immediate fixes and long-term plans. Check that the tasks fit your website, then fill in owners, due dates, and status
- If there are no evergreen page ideas, ask Codex for the top five based on customer questions. Have it add them to the workbook, with plans and drafts
Pro tip: If you skipped Part 1, ask Codex to run an audit using something like GEO Optimizer Audit first.
Suno rebuilds its models with the music industry
Suno launched v6, a family of three music models it developed alongside the Warner Music Group, BMG, and Believe, with the company saying the models were built on licensed data rather than the training set behind its older versions.
- Warner was one of three majors that sued Suno in 2024 over its training data, then settled last November in a deal promising licensed models.
- v6 includes two models for paid users and one free (v6-mini) that Suno CPO Jack Brody calls “better than any other competitor’s best paid model.”
- Round Hill, Universal, Sony, and others still have lawsuits against Suno, with an admission the company trained on YouTube also recently revealed in court.
- Fan remixes are coming next, with Suno saying artists will soon be able to opt their catalogs in and get paid.
The two AI music leaders, Suno and Udio, have both been hammered by lawsuits after early success, and both are now reworking their products with industry giants as partners. The fan remixes angle for monetization will be the interesting one to play out, potentially creating a revenue stream that incentivizes big artists to play nice.