A rule requiring insurers to spend 85% of premiums on care caps their profit margin at 15%. This creates a perverse incentive: the only way for an insurer to increase its absolute profit is to increase total healthcare spending, discouraging preventative care and cost-saving measures.
As companies deploy more AI agents, a new bottleneck emerges: managing the massive volume of exceptions and approval requests the agents generate. This creates the need for a new role, the 'Agent Supervisor,' focused on overseeing these automated workflows and handling decisions that require human nuance.
Curative deployed an AI agent, Gwen, that handles the entire sales workflow for smaller provider contracts—from research and outreach to negotiation and signing. It took the company from signing 100 contracts a week with a full team to 100 per day, with the agent alone signing more contracts in 8 weeks than the team did in a year.
Curative's credentialing department, responsible for verifying doctor licenses, was reduced to zero employees after they built an AI agent to handle the end-to-end process. The agent cut the turnaround time from 2-3 months to 12 hours and the cost from $50 to 20 cents, demonstrating AI's potential for full automation.
The future of software development will involve one senior engineer managing a team of AI agents that do the bulk of the coding. In this model, a company's spend on AI models like Anthropic could be two to five times the engineer's salary, reflecting a fundamental shift in where value is created.
Contrary to popular belief, the main obstacle for nuclear power is not engineering—safe reactors have existed since the 1960s. The real problem is a hyper-restrictive regulatory environment created in the 1980s that makes it incredibly difficult and slow to get new reactors approved and built.
The US healthcare market suffers because massive, consolidated hospital systems and payers create negotiating "loggerheads." With no viable alternatives, they reach inefficient agreements where costs are inflated for everyone. A more fragmented market with smaller players would foster more genuine and effective price competition.
Experiencing a startup failure, while devastating, provides an invaluable education in what not to do. Fred Turner states it's much easier to build a company the second time because you've already made critical mistakes in areas like hiring, process building, and firing, and learned from them.
To scale COVID testing amidst massive shortages, Curative intentionally sourced non-standard supplies (e.g., electronics swabs, different extraction materials) to avoid competing with incumbents for the same scarce resources. This 'orthogonal' approach was key to their hyper-growth from 7 to 7,000 employees in nine months.
Founder Fred Turner notes a key cultural difference: UK investors focus on credentials and mitigating worst-case scenarios, asking 'how do I not get fired?'. In contrast, Silicon Valley investors focus on the maximum possible outcome, asking 'how big could this be in ten years?', making them more open to young, unproven founders.
Signaling a potential threat to traditional SaaS, Curative canceled its Salesforce subscription after building a more effective, integrated internal CRM in just two months. This move suggests that bespoke AI-driven tools can offer superior performance and cost savings over monolithic SaaS platforms.
Instead of having LLMs parse messy files directly, a more effective method is to instruct the model to write a temporary Python script to convert the unstructured data into a standardized format. The script is then discarded. This leverages the model's strength in code generation over its weakness in direct file parsing.