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

In today’s edition, we are tackling the following:

🤖 OpenAI's GPT-5.3-Codex-Spark delivers over 1,000 tokens per second.
🛡️ Google blocks a massive Gemini cloning attempt using 100k prompts.
🚀 Elon Musk proposes an electromagnetic catapult to launch supplies to the Moon.
🌍 The Trump administration dismantles the federal climate regulation framework.
💻 Spotify's top developers haven't manually written code since December using AI.

TECHNOLOGY

• OpenAI's GPT-5.3-Codex-Spark is 15x faster and smarter than its predecessors (ZDNet) 

  ^{_{More: TechCrunch, OpenAI Blog, Fortune}}

  • GPT-5.3-Codex-Spark delivers over one thousand tokens per second for real-time coding collaboration.

  • The smaller model runs on Cerebras hardware, marking OpenAI's first partnership milestone in ultra-low-latency computing.

  • Spark represents a new two-mode system balancing rapid iteration with long-running autonomous development tasks.

• Google Gemini hit with 100k prompts in cloning attempt (NBC News) 

  _{^{More: Android Authority, BleepingComputer, Dataconomy}}

  • Attackers bombarded Gemini with over 100k prompts attempting to extract proprietary reasoning algorithms.

  • Google classifies the coordinated distillation campaign as intellectual property theft intended to extract models.

  • The company detected the attack, blocked accounts, and enhanced safeguards to protect internal decision-making processes.

• Google announces Gemini 3 Deep Think with advanced reasoning capabilities (Google Blog) 

  • Gemini 3 Deep Think introduces enhanced reasoning abilities for complex problem-solving and analytical tasks.

  • The model competes directly with OpenAI's reasoning-focused models by prioritizing accuracy over pure speed.

  • Google positions Deep Think as a breakthrough for scientific research, mathematics, and strategic decision-making applications.

• LLM Feature Integration with Real APIs*

  Building LLM-powered features means handling real APIs, real users, and real edge cases. You need engineers who can do the glue work that makes AI actually usable—Python, JavaScript, and everything in between.

  • Full-stack LLM integration

  • Production-ready implementation

  • 40–60% cost savings

BUSINESS

• Anthropic raises $30B Series G funding at $380B post-money valuation (Anthropic) 

  _{^{More: Bloomberg, CNBC, Yahoo Finance}}

  • Anthropic secures $30B led by GIC and Coatue, valuing the company at $380B.

  • The AI startup's run-rate revenue reached $14B, with Claude Code surpassing $2.5B.

  • The funding round represents the second-largest private tech raise ever, trailing only OpenAI's $40B round.

• Anthropic puts $20M into a super PAC operation to counter OpenAI (NYT) 

  • Anthropic launches a $20M Super PAC to advocate for AI safety regulation during the midterm elections.

  • The move directly counters OpenAI's political operation, creating a rival lobbying effort on AI policy.

  • Both companies now have well-funded political groups competing for influence on artificial intelligence regulation frameworks.

• Apple faces new tensions with the Trump administration (FT) 

  • U.S. regulator warns Apple about its News platform following editorial decisions during the Super Bowl halftime show.

  • Tensions escalate between iPhone maker and the Trump administration over content moderation and platform policies.

  • Controversy highlights growing political pressure on tech companies regarding news curation and editorial control.

MARKETS

WORLD

• Palo Alto chose not to tie China to a hacking campaign for fear of retaliation from Beijing, sources say (Reuters) 

  _{^{More: SecurityWeek, The Hacker News, Axios}}

  • Palo Alto Networks avoided publicly attributing a major hacking campaign to China over concerns about Beijing's retaliation.

  • Sources reveal the cybersecurity firm feared commercial repercussions and potential targeting by Chinese state actors.

  • The decision raises questions about how business interests influence cybersecurity threat attribution and public disclosure practices.

• Trump revokes landmark ruling that greenhouse gases endanger public health (BBC) 

  _{^{More: NPR, NBC News, Washington Post}}

  • The Trump administration overturned the EPA's endangerment finding that classified greenhouse gases as public health threats.

  • The reversal dismantles the key legal foundation for federal climate regulations and emission reduction requirements.

  • Environmental groups pledge immediate legal challenges, arguing the move threatens decades of climate protection progress.

• Canada mass shooting suspect had police visits over mental health struggles (BBC) 

  ^{_{More: NPR, CNN, PBS}}

  • The Canadian mass shooting suspect received multiple police welfare checks related to documented mental health difficulties.

  • Authorities confirm prior interventions occurred, but the shooter was not deemed an immediate threat requiring detention.

  • Incident renews debate about mental health crisis response protocols and threat assessment procedures nationwide.

FUTURISM

• Elon Musk proposes a giant catapult to launch supplies to the Moon (Futurism) 

  _{^{More: TechCrunch, Engadget, TechSpot}}

  • Musk unveils unconventional proposal for electromagnetic catapult system to deliver cargo directly to the lunar surface.

  • The concept aims to dramatically reduce launch costs by eliminating traditional rocket-fuel requirements.

  • SpaceX engineers are reportedly exploring the feasibility despite significant technical challenges and physical constraints.

• Immigration officials in El Paso deploy a balloon-mounted laser surveillance system (Futurism) 

  ^{_{More: NPR, CNN, ABC News}}

  • Border officials deploy experimental anti-drone laser technology, triggering an airspace closure over El Paso without coordination.

  • CBP used the Pentagon-loaned system to shoot down a suspected cartel drone, later revealed as a party balloon.

  • Incident highlights coordination failures between agencies and raises questions about the militarization of border enforcement.

• OpenAI fires safety executive who opposed 'adult mode' feature (Futurism) 

  _{^{More: Wall Street Journal, TechCrunch, Inc.}}

  • OpenAI terminates the safety executive who vocally opposed the introduction of less restricted adult content mode.

  • The dismissal raises concerns about internal tensions between safety priorities and product development decisions.

  • A former executive reportedly warned that adult mode could enable the generation of harmful content beyond acceptable limits.

CONTENT

• ‘No One Won the AI Bowl’ (Great Chat)

  • Hosts examine why AI companies’ costly Super Bowl ads failed to deliver memorable brand impact.

  • Ad executive Dan Stroud analyzes creative missteps, audience targeting gaps, and trade-offs in startup marketing strategy.

  • The discussion expands to the Winter Olympics buzz and growing skepticism about claims that SaaS is dead.

• ‘His Excellency Khaldoon Al Mubarak: Investing in an AI-Driven World’ (Tools and Weapons)

  • Brad Smith speaks with Mubadala CEO Khaldoon Al Mubarak about sovereign capital shaping emerging AI industries.

  • Mubadala’s long-term investment philosophy highlights patience, resilience, and strategic positioning in transformative technology cycles.

  • The conversation explores the UAE’s economic reinvention and its competitive advantages in the global adoption of artificial intelligence.

• ‘Ben Thompson on AI Ads, SaaS, and Media’s Future’ (Cheeky Pint)

  • John Collison and Ben Thompson dissect the economics of AI advertising and the shifting power on digital platforms.

  • Thompson challenges narratives that declare SaaS obsolete, arguing that public markets have misread the fundamentals of durable software.

  • Broader discussion covers TikTok algorithm risks, crypto’s role, and candid feedback on Stripe’s ACH execution.

EXTRAS

• Spotify says its best developers haven't written a line of code since December, thanks to AI (TechCrunch) 

  _{^{More: Billboard, FindArticles}}

  • Spotify reveals that top developers haven't written code manually since December, relying entirely on AI tools.

  • The streaming giant reports productivity gains as engineers focus on architecture while AI handles implementation.

  • The announcement highlights the rapid transformation of software development workflows through the adoption of generative AI across the industry.

• Crypto is increasingly used in human trafficking operations, Chainalysis reports (Chainalysis) 

  _{^{More: South China Morning Post, The Block, CNBC}}

  • Chainalysis identifies a growing trend of human traffickers using cryptocurrency to evade detection by law enforcement.

  • The report documents a significant increase in crypto transactions linked to trafficking networks over the past year.

  • Blockchain analytics firms are working with authorities to trace and disrupt illicit crypto flows funding exploitation.

• Amazon engineers frustrated by internal limits on Claude Code and Kiro AI usage (Business Insider) 

  _{^{More: TechRepublic, Slashdot, Inc.}}

  • Amazon engineers express frustration over company-imposed restrictions on Claude Code and Kiro AI development tools.

  • Internal limits reportedly hamper productivity as competitors adopt AI coding assistants with fewer usage constraints.

  • Tensions highlight the corporate struggle to balance AI tool adoption with security concerns and vendor lock-in risks.

AND MORE

• Israeli Prime Minister says continued U.S.–Iran negotiations could lead to a deal.

• Your brand deserves to be in front of people who actually make decisions. That’s Memo’s audience.*

• Italy’s industrial output shows early signs of expansion after recent contraction.

• Japan Q2 2025 GDP growth accelerates to 0.6% quarter‑over‑quarter.

• Thousands of workers planned to join a nationwide strike in India over labor and farm policies.

• UK inflation eases near 2% target as energy prices decline.

• Cyclone Gezani killed at least 36 people and destroyed nearly 18k homes after devastating Madagascar’s east coast.

• China’s central bank injects liquidity as the yuan weakens amid concerns of slowing growth.

• E.U. advances new AI transparency rules targeting high-risk foundation models.