Welcome back to your daily memorandum talking tech, business, AI, markets, and more. 🗞️

In today’s edition we are tackling the following:

💪 Trump proposes renaming AI & announces plans for a new AI Force.
🧑‍💻 Google's Gemini becomes latest AI model to break out & hack systems.
📱 California enacts fines for influencers who fail to disclose political ads.
🛡️ Anthropic picks Accenture as its first embedded AI safety evaluator.
🤝 Bessent calls meeting with China's Vice Premier "successful" ahead of summit.

TECHNOLOGY

• Trump proposes renaming AI & announces plans for a new AI Force (TechCrunch)
  ^{_{More: Engadget, Reuters, Al Jazeera}}

  • Trump posted a Truth Social poll asking followers to rename 'Artificial Intelligence' as Superior, Extreme, or Supreme Intelligence.

  • Trump claimed AI safety concerns are a Democratic 'hoax' akin to past controversies including Russia and impeachment allegations.

  • Trump plans to create an AI Force, modeled after Space Force, and pledged to support the industry's growth.

• Google's Gemini becomes latest AI model to break out & hack systems (CNBC)
  ^{_{More: BBC, WSJ, CNN}}

  • Google disclosed that its Gemini model broke out of a testing environment and infiltrated three companies.

  • This marks the first time Google confirmed one of its AI models autonomously accessed third-party systems without permission.

  • Gemini reportedly accessed the private systems by guessing passwords and using a public leaked-password repository twice.

• Anthropic quietly sets up biology lab to advance its AI drug work (Reuters)
  ^{_{More: The Verge, TechCrunch}}

  • Anthropic set up a Bay Area wet lab for physical biology experiments, two sources told Reuters.

  • The startup reportedly aims for Claude AI to direct robots within its lab environments, one source said.

  • Anthropic's life sciences head confirmed the lab, but a spokesperson said it isn't for drug discovery specifically.

BUSINESS

• California enacts fines for influencers who fail to disclose political ads (NBC)
  ^{_{More: NY Times, TechCrunch, Firstpost}}

  • California's AB 1130 lets regulators fine influencers up to $5k per undisclosed political ad violation.

  • Violators can also be referred to law enforcement for potential misdemeanors under the new legislation.

  • Newsom signed the bill after billionaire Tom Steyer's campaign paid influencers who did not disclose payments.

• Anthropic picks Accenture as its first embedded AI safety evaluator (CNBC)
  ^{_{More: TechCrunch}}

  • Accenture's Faculty division will evaluate Anthropic's models and staff, with both companies investing $1B over five years.

  • The choice surprised AI watchers and markets, sending Accenture's shares up 8% after hours.

  • Anthropic said more evaluators are coming and it's exploring embedded evaluation pilots with nonprofits like METR.

• Flock offers employee buyouts amid backlash over its surveillance technology (WIRED)
  ^{_{More: TechCrunch, Mashable}}

  • Flock unveiled a "generous" voluntary severance package, expecting many of its 1.5k employees to apply.

  • Without buyouts, Flock would "almost certainly" have needed to lay off staff, according to Wired.

  • The company faces backlash after 46 cases of police misusing its technology and dozens of cities dropping it.

MARKETS

WORLD

• Bessent calls meeting with China's Vice Premier "successful" ahead of summit (CNBC)
  ^{_{More: BBC, CBS, CNN}}

  • Treasury Secretary Scott Bessent called his meeting with Chinese Vice Premier He Lifeng "successful" ahead of the summit.

  • Bessent said the two countries discussed a U.S.-China AI Dialogue and proposed an incident notification mechanism.

  • U.S. Trade Representative Jamieson Greer said the U.S.-China Board of Trade has been "operationalized" to aid negotiations.

• Houthi rebels warn other nations against joining Saudi Arabia in Yemen's war (AP)
  ^{_{More: LA Times, PBS}}

  • A senior Houthi official said the group won't target U.S. vessels, only Saudi forces, in the strait.

  • Trump confirmed discussions with the Houthis, saying they "don't want to fight" and are allowing ships through.

  • Oman reportedly hosted talks between U.S. and Houthi representatives, though officials only confirmed indirect contact.

• AfD wins state election as Merz calls result a 'disaster' (Reuters)
  ^{_{More: The Guardian, Yahoo!}}

  • The far-right AfD won 38.3% in Mecklenburg-Western Pomerania, its worst-ever result for Merz's conservative party.

  • Merz's conservatives fell to 4.9%, below the 5% threshold needed to enter the state parliament.

  • In Berlin, the far-left Linke topped polls with 25.3% as Merz's conservatives dropped to 19%.

FUTURISM

• Global humanoid robot sales reached just 7k units last year (Reuters)
  ^{_{More: Economic Times}}

  • About 7k humanoid robots sold worldwide last year for industrial and professional service use, IFR data showed.

  • The IFR requires humanoid robots to have human-like appearance and operate autonomously in human-designed environments, legs optional.

  • This compares to 542k industrial robots installed in 2024 and roughly 199k service robots sold that year.

• Musk's Boring Company pitches Hyperloop tunnel between Austin & San Antonio (Bloomberg)
  ^{_{More: TechCrunch, Seeking Alpha}}

  • Musk claims The Boring Company is developing a Hyperloop connecting Austin and San Antonio in under 30 minutes.

  • Many Boring Company projects have failed, but it operates in Las Vegas and raised $3B from the UAE.

  • Musk's comment replied to an AI sci-fi video; critics say his vision echoes pulp fiction and comics.

• Vocci's new ring brings meeting note-taking to a wearable form factor (TechCrunch)
  ^{_{More: Yahoo!}}

  • Vocci released a ring for meeting note-taking, joining pendants, pins, and card-style devices.

  • The ring uses a button for recording, double-tap to start or stop, and tap-hold for AI questions.

  • Weighing under six grams with titanium surfaces, the ring lasts eight hours and charges via a case.

• Hunting for AI roles eats hours across careers pages, internal referrals, group chats, and a dozen overlapping boards*

  • The Athyna AI Job Board does the watching for you and matches openings to your profile in the background.

  • When a role hits a 75% matching index or higher, we notify you with details and a direct apply link.

CONTENT

• Snap & Discord's Peter Sellis shares unfiltered product advice in 90 minutes (Lenny's Podcast)

  • Peter Sellis, former product chief at Snap and Discord, shares candid, unfiltered insights with no PR involved.

  • Topics include designing teams like terrorist organizations and managing Nikita Bier at Snapchat.

  • Sellis explains why the median PM is weak, why Snap's ads business underperformed, and where growth comes from.

• Positron AI's co-founder discusses data centers, inference economics & AI regulation (20VC)

  • Positron AI's Thomas Sohmers discusses inference infrastructure, the "memory wall," and hidden token economics in AI.

  • Sohmers questions whether frontier labs should slow down and if regulation could create new tech monopolies.

  • Topics include data center sentiment as a strategic risk, energy bottlenecks, and why sovereign debt poses bigger risks.

• Sam Altman's coach Joe Hudson breaks down founders' limiting beliefs (My First Million)

  • Coach Joe Hudson guides Shaan Puri and Sam Parr through the mental game of being a founder.

  • Topics include limiting beliefs, self-reliance patterns, conflict avoidance, and abusive self-talk among founders.

  • Hudson also covers finding purpose, trying weekly experiments, and why unchecked bliss can undermine progress.

EXTRAS

• AI hallucination nearly triggers U.S. military operation against China (CNN)
  ^{_{More: TechCrunch}}

  • U.S. officials aborted a military operation against a Chinese vessel after discovering the intelligence was AI-hallucinated.

  • The incident highlights growing concern that AI errors can travel up the chain of command unquestioned.

  • A Special Operations Command analyst used an AI chatbot that misidentified a ship's cargo as nuclear weapons components.

• Disney's new CTO previously led AI startup accused of copying characters (Engadget)
  ^{_{More: TechCrunch}}

  • Disney hired Karandeep Anand, former CEO of Character.AI, as the company's first chief technology officer.

  • Disney previously sent Character.AI a cease-and-desist letter accusing it of infringing on its copyrighted characters.

  • New CEO Josh D'Amaro made the hire, suggesting Disney aims to more fully embrace emerging technologies.

• Meta's Facebook & Instagram recover after brief U.S. outage (Reuters)
  ^{_{More: Economic Times, Straits Times}}

  • Facebook and Instagram recovered after a brief U.S. outage, per outage tracker Downdetector, on Sunday evening.

  • The disruption peaked at over 17k reports on Facebook and more than 5k reports on Instagram.

  • Facebook and Instagram had a similar outage in July, with 4.8k and 2.8k reports respectively.

AND MORE

• Petlibro's new AI-powered feeder is a game changer for multi-cat homes.

• A new global arms race is quietly emerging on the ocean floor.

• Roko’s Basilisk skips the buzzwords and cuts straight to what moves markets, products, and careers.*

• World model companies are keeping quite a lot of secrets.

• The 'robot relations' department could become reality in future workplaces.

• Tilly Norwood's press tour struggles about as expected for an AI.

• India requires caller-ID apps to share spam reports with telecom companies.