Stanford’s 2025 AI Index Report: Insights

Also, Shopify’s leaked internal memo re: AI usage is ruffling feathers

Author

Jack Lajoie
April 08, 2025 • read time ~ 7 minutes

⚡️ Headlines

🤖 AI

White House Orders Federal Agencies to Appoint Chief AI Officers and Develop AI Strategies - The White House mandates federal agencies to designate chief AI officers and formulate strategies to expand AI use, rescinding prior safeguards. [Reuters]

OpenAI Explores Acquisition of Jony Ive and Sam Altman's AI Device Startup - OpenAI has engaged in discussions to acquire the AI device startup founded by former Apple designer Jony Ive and OpenAI CEO Sam Altman. [The Information]

Meta Executive Denies Claims of Inflated Llama 4 Benchmark Scores - Meta's VP of Generative AI refutes allegations that the company manipulated Llama 4's benchmark results. [TechCrunch]

Google's AI-Powered Search Now Answers Image-Based Queries - Google's AI Mode Search feature can now interpret and respond to questions related to images. [Ars Technica]

Microsoft AI Chief Advocates for Slightly Behind AI Models - Microsoft's AI leader suggests that AI models trailing the latest by a few months may offer advantages in stability and reliability. [CNBC]

Stanford Study Highlights China's Rapid Progress in AI Research - A Stanford report reveals China's swift advancements in AI, nearing the United States in research capabilities. [Wired]

AI Integration in Radiology Departments on the Rise - Approximately two-thirds of U.S. radiology departments now utilize AI in some capacity, according to a recent survey. [The Washington Post]

🤳 Social Media

Snapchat Launches Sponsored AI Lenses for Brands - Snapchat introduces Sponsored AI Lenses, allowing brands to engage users with AI-generated interactive experiences. [TechCrunch]

Instagram Tests Monthly Performance Recap Feature - Instagram is piloting a new feature providing users with a monthly summary of their content performance. [Social Media Today]

🔬 Research

Multi-Agent Systems Enhance Clinical Decision Support - A study proposes a multi-agent system architecture to improve clinical decision support through ethical AI governance. [arXiv]

🎱 Random

Apple Expedites iPhone Shipments to Avoid Tariffs - Apple rapidly transported five flights of iPhones from India and China to the U.S. within three days to circumvent impending tariffs. [The Times of India]

Google Allegedly Pays AI Staff to Remain Inactive to Prevent Rival Employment - Reports suggest Google compensates certain AI employees to refrain from working elsewhere for up to a year. [TechCrunch]

🔌 Plug-Into-This

Stanford's 2025 AI Index Report offers a sweeping analysis of global AI development and its implications for enterprise strategy, spotlighting accelerating capabilities, declining costs, and increasing—but not always profitable—adoption rates. The report serves as a quantitative backbone for strategic planning amid AI’s rapid maturation.

  • The United States led AI development in 2024, producing 40 notable models, far outpacing China and Europe.

  • Training compute is doubling roughly every five months, with dataset sizes following a similar growth trajectory.

  • Inference costs have dropped dramatically—from $20 to $0.07 per million tokens between late 2022 and 2024.

  • Private AI investment jumped 26% year-over-year, reaching $252.3 billion globally.

  • AI adoption among organizations climbed to 78%, but most report revenue impacts below 5%.

🧠 In simplest terms, AI is getting faster and cheaper, but businesses are still figuring out how to make real money from it. The rapid democratization of AI capabilities underscores the need for enterprises to not only adopt AI but to strategically integrate it in ways that yield tangible business value.

Shopify CEO Tobi Lütke has instituted a policy requiring teams to prove that a task cannot be handled by AI before requesting additional staffing. The move formalizes AI usage as a core operational competency across the company.

  • AI literacy is now embedded in performance reviews and peer assessments.

  • Employees are expected to treat AI tools as essential teammates in daily workflows.

  • Lütke described this approach as a reflexive expectation—not a stretch goal—for modern teams.

  • The policy supports Shopify’s broader pivot toward enterprise clients and operational efficiency.

  • The company joins others in tightening headcount while deepening AI adoption.

🤖 Basically, Shopify wants employees to use AI first and hire people only if AI can't do the job. Fair enough, you'd hate to see a company hiring a person to draw spreadsheets by hand when they could be using Microsoft Excel…but the announcement feels rather crass in this current social and political climate.

Waymo is exploring the use of interior vehicle camera footage—including potentially identifiable rider data—to train generative AI models and support ad-related services. Riders will be offered the ability to opt out of this data usage.

  • A draft privacy policy suggests footage could be used to improve AI and personalize services.

  • Opt-out mechanisms are planned but not yet finalized in terms of notification or deployment.

  • Waymo currently uses personal data for safety, cleanliness, and service optimization.

  • It pledges not to share personal data with other Alphabet entities without consent.

  • The approach raises fresh questions about consent, surveillance, and commercialization in autonomous transport.

🚘 As is usually the case with tech companies, they'll take everything they can and more, but you do have an opt out buried somewhere deep in the terms and conditions…hidden behind a few levels of abstraction, most likely.

 🆕 Updates

📽️ Daily Demo

🗣️ Discourse