Anthropic beats OpenAI, Attenborough slams AI clones, Wall Street’s AI push

Welcome back to a new edition of AI Odyssey!

Elon Musk is suing Microsoft, accusing the tech giant of colluding with OpenAI and Sam Altman to "loot" the nonprofit-turned-for-profit AI startup. Musk claims Microsoft pushed OpenAI toward profit-driven secrecy and added board member Reid Hoffman as a co-defendant in the case.

^{Sponsored by HubSpot}

[Webinar] From Insights to Growth: Using Product Data to Fuel Smarter Marketing

Join us on December 4th to learn how to turn product insights into impactful marketing with HubSpot, Amplitude, and Aptitude 8. Join us to learn:

• How AI and data are reshaping the marketing landscape and how to stay ahead

• The power of PQLs (product qualified leads) and how they can fuel faster growth

• Actionable tips on aligning marketing and product teams to harness data effectively

Register for Webinar

1: Anthropic outperforms OpenAI in AI research test

The news: In a groundbreaking evaluation, Anthropic’s Claude Sonnet 3.5 outperformed OpenAI’s o1-preview in solving complex AI research problems, marking a significant milestone in the development of AI capable of improving itself.

Key results:

• Claude excelled in five tests, winning decisively in two.

• OpenAI’s o1-preview dominated two tests.

• Both lagged behind top human researchers, who scored twice as high.

Why it matters: These tests, by METR, assessed how AI handles tasks like forming hypotheses and running experiments—crucial steps for AI to improve itself.

The big picture:

• Progress: Claude matched average human researchers in two tests, hinting at AI’s growing capabilities.

• Challenges: Tasks like creating language models without division pushed both AI and humans to their limits.

What’s next:

• Safety concerns: Governments are monitoring self-improving AI, citing potential risks.

• Industry stakes: Recursive AI could transform the field but raises serious ethical and safety questions.

While AI hasn’t surpassed humans yet, its steady progress makes a self-improving future increasingly plausible.

2: David Attenborough condemns AI voice replica: ‘Profoundly disturbed’

The news: Sir David Attenborough has expressed outrage after learning that AI-generated replicas of his iconic voice are being used without his consent. The 98-year-old broadcaster criticized the technology for stealing his identity, saying, “I am profoundly disturbed to find that my identity is being stolen by others and greatly object to them using it to say whatever they wish.”

Why it matters: Attenborough’s voice, synonymous with truth and authority, is globally recognized. The misuse of AI to replicate trusted figures raises ethical concerns about misinformation, particularly in matters like war and politics, as highlighted by AI researcher Dr. Jennifer Williams.

The big picture:

• AI voice cloning is becoming alarmingly accurate, as demonstrated in a BBC News segment comparing real and fake Attenborough recordings with indistinguishable results.

• The issue underscores broader fears about deepfake technology being weaponized to spread false narratives, eroding public trust in authentic voices.

Looking ahead: As AI-generated content grows more sophisticated, calls for stricter regulation and ethical safeguards will intensify. For now, figures like Attenborough serve as powerful reminders of the risks posed by unregulated AI advancements.

3: Wall Street banks bet big on AI infrastructure boom

The news: Wall Street giants like Morgan Stanley and Blackstone are diving into the $1 trillion AI infrastructure race, funding data centers, power grids, and telecom networks to support AI’s explosive growth.

Why it matters: AI’s transformative potential demands massive investment, creating lucrative opportunities for banks and private equity—but also significant risks.

Key players:

• Tech giants like Google and Microsoft spent $52.9 billion last quarter alone on AI infrastructure.

• Banks and private equity are funding “AI factories,” with creative deals like using Nvidia chips as loan collateral.

Challenges: Energy demands, speculative risks, and uncertain returns loom over this unprecedented spending spree.

The takeaway: Wall Street sees AI infrastructure as a long-term goldmine, betting on tech giants’ deep pockets to keep the boom alive.

4: Apple’s AI ambitions target the smart home

The news: Apple is gearing up for a major push into the smart home market, leveraging AI to deliver new devices and features. Reports suggest a wall-mounted smart hub will debut next year, followed by Apple-branded security cameras in 2026.

Why it matters:

• Second chance for smart homes: AI could finally unlock the potential of smart homes, moving beyond the clunky setups of the past to devices that understand context and deliver useful responses.

• Apple's edge: With its privacy reputation and seamless integration, Apple could stand out from competitors like Amazon and Google.

Challenges ahead:

• Proving AI’s value: Apple needs to show how AI can meaningfully improve daily life, like automating app functions through its new App Intents feature.

• Monetization: Turning AI-powered smart devices into a profitable business model remains unclear, especially as Apple warns its new products may not match the iPhone’s profitability.

The big picture: Success hinges on Apple’s ability to create truly useful AI-driven services and convince users to pay for them. The smart home may be its next frontier, but its long-term success in AI will depend on more than just hardware.

AI NOTES 🗒️

Extra read: Coca Cola’s AI-generated ad controversy, explained (Forbes)

Building AI: Inside Microsoft's struggles with Copilot (Business Insider)