AI in Marketing

The Social Contract Is the Real AI Story

On a recent episode of The Artificial Intelligence Show, host Paul Roetzer raised a question that most AI coverage skips past: what happens to the social contract when AI changes who does the work?

We talk about the social contract like it's a firm deal. You build skills, work hard, get wages and stability in return. Employers provide opportunity. Society provides education and safety nets. In practice, that arrangement has been fraying for decades. Pensions disappeared. Real wages flattened. Entire industries moved overseas with minimal political consequence. AI didn't break the social contract. It's just the next thing testing what's left of it.

If machines handle more of the tasks that make up knowledge work, the question isn't just who loses a job. It's who captures the value. Do productivity gains flow to workers through better pay or shorter hours? Or do they concentrate among company owners and a small technical class? History suggests the answer, and it's not the optimistic one.

Roetzer pointed to a conversation between Peter Diamandis and Andrew Yang, who has spent years on this problem. Yang raised the idea of universal basic services: instead of just handing people cash, guarantee healthcare, education, and retirement security for a modest monthly cost. It's one idea. There aren't many others getting serious attention.

That's the problem. The technology is moving. The policy conversation is barely started. The layoffs won't be announced as layoffs. They'll be called efficiency gains, workforce optimization, strategic realignment. By the time the pattern is obvious, the old arrangement will already be gone.

There is no federal commission studying AI labor displacement. No proposed legislation addressing how productivity gains get distributed. No serious institutional response to any of it. The gap between what AI can do and what the political system is prepared to handle is the real story.

Next.

Microsoft has taken legal action to challenge the Pentagon's decision to blacklist Anthropic, a move that could have significant implications for the tech industry. The company is seeking a temporary restraining order to prevent the Department of Defense from enforcing its supply chain risk designation against Anthropic, an AI startup whose models are integrated into Microsoft's business products. This legal battle highlights the growing tension between tech companies and government agencies over security designations that can impact commercial AI partnerships. The Pentagon's decision to blacklist Anthropic, citing unspecified security concerns, threatens Microsoft's enterprise AI strategy and raises questions about the fairness and transparency of government security reviews. The case underscores the challenges faced by the government in regulating AI while maintaining security, as traditional risk assessment frameworks struggle to keep pace with the rapidly evolving AI landscape. Microsoft's public challenge to the Pentagon indicates the seriousness of the perceived threat to its AI investments, and the outcome could set a precedent for how government security reviews affect the commercial AI sector.

On a different note.

Yann LeCun, former chief AI scientist at Meta, has launched a startup named Advanced Machine Intelligence (AMI) in Paris, securing over $1 billion in funding to develop AI systems that understand the physical world. LeCun believes that true human-level AI will emerge from mastering the physical world rather than language, challenging the prevailing focus on large language models (LLMs) by major AI labs like OpenAI and Meta. AMI, valued at $3.5 billion, is backed by investors including Cathay Innovation, Greycroft, and Bezos Expeditions, with notable figures like Mark Cuban and Eric Schmidt also supporting the venture. The startup plans to create AI systems with persistent memory and reasoning capabilities, aiming to collaborate with industries such as manufacturing and robotics. LeCun, a Turing Award winner, has been critical of the limitations of LLMs, arguing they won't achieve human-level intelligence. AMI's leadership includes former Meta colleagues and industry experts, with offices in Paris, Montreal, Singapore, and New York. The company intends to build open-source technology, emphasizing that AI should not be controlled by a single entity. AMI's initial projects will involve partnerships with companies like Toyota and Samsung, with the goal of developing a universal world model for broader applications.

Meanwhile.

Meta Platforms has acquired Moltbook, a social network designed for AI agents, integrating its founders Matt Schlicht and Ben Parr into Meta Superintelligence Labs. This acquisition highlights the competitive drive among tech companies to secure AI expertise as autonomous agents gain prominence. Moltbook, initially a niche project, has sparked discussions on the proximity of AI to human-like intelligence. Despite OpenAI CEO Sam Altman suggesting Moltbook might be a fad, he acknowledged the potential of its technology. Moltbook's development, primarily through Schlicht's AI assistant, faced security issues that were resolved after being flagged by cybersecurity firm Wiz. Financial details of the acquisition remain undisclosed, and Schlicht and Parr are set to join Meta on March 16.

In other news.

Artificial intelligence is advancing drug discovery for diseases like Parkinson's, antibiotic-resistant infections, and rare conditions. AI accelerates the identification of new compounds by analyzing vast chemical libraries, as demonstrated by James Collins at MIT, who found potential treatments for gonorrhea and MRSA. These AI-generated compounds target bacteria differently from existing antibiotics, offering hope against drug resistance. In Parkinson's research, Michele Vendruscolo at the University of Cambridge uses AI to identify compounds that may prevent the disease by stabilizing proteins before they form harmful clumps. This approach could lead to more effective treatments. Additionally, existing drugs are being repurposed for rare diseases, as shown by David Fajgenbaum's personal experience with Castleman disease.

After that.

TechCrunch reports that Amazon is expanding its healthcare AI assistant, Health AI, to its website and app. Previously exclusive to the One Medical app, Health AI offers services like answering health questions, managing prescriptions, and booking appointments. Users do not need to be Prime subscribers or One Medical members to access it. While Health AI can provide general health information without accessing personal medical data, it can also offer personalized guidance and connect users with healthcare professionals. Amazon assures that interactions occur in a HIPAA-compliant environment, with encryption and strict access controls, though specifics on encryption remain undisclosed. Health AI can access user health information via the Health Information Exchange, allowing it to interpret lab results and medical records. Prime members in the U.S. receive up to five free consultations with One Medical providers for common conditions, while non-Prime users can use a pay-per-visit option. Users can sign up on the Amazon Health page.

Finally.

Drug pricing decisions in the pharmaceutical industry significantly impact a company's reputation, as highlighted by Carreen Winters of MikeWorldWide. She emphasizes that pricing is not just a financial issue but a reputational one, influencing public perception. For instance, Novo Nordisk's recent 50% price cut on GLP-1 drugs reflects public concerns about access and affordability during economic challenges. Winters advises that communications leaders should be involved in pricing strategy discussions to anticipate public reactions and align decisions with the company's values. Companies should consider how pricing aligns with their public image and stakeholder expectations. Without proactive communication, pricing decisions risk being perceived as prioritizing profits over patients, potentially damaging trust and reputation.