Businesses are finding practical uses for AI, while stressing that people must remain accountable. The Baldwin Group says its Catalyst program completed more than 47,000 AI-assisted tasks in 17 weeks, standardized 27 processes and achieved quality rates above 98%. The company says the time saved should support client advice. In one claims review, AI identified coverage gaps that contributed to an additional $71,000 payment, with a professional retaining responsibility for the decision.
Iowa executives described similar support for office and project work. Baker Group President Joel Veenstra says AI helps with contract review, estimating and document coordination, and the company is investing in data, governance, employee skills and flexible technology choices. McGough Construction’s Mike Harryman sees potential in quantity analysis, scheduling and cost forecasting, freeing teams to spend more time with clients and project partners. Yet companies broadly still struggle to fit AI into workflows: it can process data quickly but make mistakes and miss nuances in communication.
That need for oversight also features in advice on AI agents. University of San Francisco adjunct professor Shomit Ghose warns that agents can misread instructions or use inaccurate information; a request for the cheapest flight, for instance, might produce an inconvenient itinerary. He advises people to review travel bookings, retain control of major financial actions and restrict access to sensitive records. A November 13, 2026 opinion essay takes a wider view, arguing that AI reliance can weaken students’ learning, workers’ skills and human judgment.
Debate over AI’s economics and global competition continues. Shanghai-based economist Andy Xie compares the U.S. AI boom to a Ponzi scheme, arguing that data-center growth helps companies raise debt and that investor optimism can fuel both rapid development and waste. AI entrepreneur Kai-Fu Lee, meanwhile, says Chinese developers have adapted to U.S. chip export limits by building capable models with fewer resources. He estimates some Chinese open-weight models trail leading U.S. systems from OpenAI and Anthropic by six to 12 months, and says market competition—not a coordinated government plan—has driven the shift toward open models.
In finance, AlphaPai announced its international launch at Hong Kong FinTech Week × StartmeupHK 2026 on November 13. The company says its China-market products serve more than 120,000 professional users across over 8,000 institutions and asset managers. Its PaiWork research workstation and PaiMeeting assistant support information gathering, meeting review and research, including tools for company analysis, management questions and earnings-estimate changes across Asian markets.
Two other developments highlight AI’s reach and limitations. A Skåne University Hospital study of GE Healthcare’s Venue GO ultrasound system found poor agreement with expert physicians: the software overcounted lung B-lines in a study of 55 hospitalized patients, with 574 of 672 collected clips suitable for analysis. Shield AI opened a Bengaluru office, its second in India, and plans to hire engineers to adapt its Hivemind autonomy software for local customers. The move follows a Nevada test in October in which two combat drones performed coordinated maneuvers beyond visual range. JSW Defence is building a V-BAT production facility in Hyderabad as part of a $90 million investment; India selected V-BATs and Hivemind licenses for its Army in January.
Key Takeaways
- The Baldwin Group reports more than 47,000 AI-assisted tasks in 17 weeks, quality rates above 98% and an AI-supported claims review that contributed to an additional $71,000 payment.
- Baker Group is using AI for contract review, estimating and document coordination while investing in data, governance and employee skills; McGough sees uses in project analysis, scheduling and cost forecasting.
- Companies still face challenges integrating AI into workflows, and experts urge human review of AI agents, especially for financial decisions, travel bookings and sensitive personal data.
- A November 13, 2026 opinion essay argues that AI reliance can weaken learning, professional skills and human judgment.
- Economist Andy Xie says the U.S. AI and data-center boom carries bubble-like risks, including debt-fueled expansion and waste.
- Kai-Fu Lee estimates some Chinese open-weight models are six to 12 months behind leading U.S. models from OpenAI and Anthropic, and attributes China’s move toward open models to market competition.
- AlphaPai launched internationally with PaiWork and PaiMeeting; it says its China-market products serve over 120,000 professionals across more than 8,000 institutions and asset managers.
- A Swedish hospital study found GE Healthcare’s Venue GO AI ultrasound tool overcounted pulmonary B-lines compared with expert physicians.
- Shield AI opened a Bengaluru office to hire engineers and adapt Hivemind for India; JSW Defence’s V-BAT facility in Hyderabad is part of a $90 million investment.
Baldwin Group says AI will make insurance advisors more valuable
The Baldwin Group says AI will change insurance brokerage by speeding up tasks such as gathering information and comparing coverage. Its Catalyst program completed more than 47,000 AI-assisted tasks in 17 weeks, with quality rates above 98 percent, while standardizing 27 processes. The company says the time saved should go toward advising clients, not just cutting costs. In one claims review, AI helped identify coverage gaps that led to an additional $71,000 payment, while a professional remained responsible for decisions.
Iowa executives describe how AI may reshape business
Iowa business leaders shared their views on AI during the Business Record’s final Power Breakfast of 2026 and in follow-up comments. Baker Group President Joel Veenstra said AI supports office tasks such as contract review, estimating, and document coordination, helping teams save time and reduce rework. Baker Group is investing in data, governance, employee skills, and flexible technology choices while keeping people accountable. McGough Construction General Manager Mike Harryman said AI could speed up tasks such as quantity analysis, scheduling, and cost forecasting, leaving more time for teams to work with clients and project partners.
Essay argues AI can weaken human skills and judgment
In an opinion essay published November 13, 2026, the writer argues that AI threatens human flourishing and should be resisted. The essay says students learn less when AI replaces their own work, and workers who rely on AI may become supervisors of a machine rather than skilled practitioners. It also argues that using AI to replace skills can reduce people’s ability to think seriously and can undervalue human expertise. The author presents these claims as a case against relying on AI.
Experts say keep humans in control of AI agent decisions
AI agents can handle many tasks, but University of San Francisco adjunct professor Shomit Ghose warns that they may misunderstand instructions or rely on inaccurate information. A request for the cheapest flight, for example, could lead to a very early departure or several layovers unless the user sets clear limits. Ghose says people should keep control of major financial actions, such as moving money or making trades, and should review flight and hotel bookings before they are made. He also advises limiting agents’ access to personal information, such as bank accounts and medical records.
Economist compares America’s AI boom to a Ponzi scheme
Shanghai-based independent economist Dr Andy Xie argues that America’s AI boom is beginning to resemble a Ponzi scheme. He says AI companies increasingly depend on a data center boom that helps them raise more money through debt markets. Xie also describes market bubbles as products of excessive investor optimism, which can speed up technology development but create waste. He questions whether informed market insiders truly believe AI firms are worth trillions of dollars.
Kai-Fu Lee challenges three myths about China’s AI industry
AI entrepreneur Kai-Fu Lee says U.S. chip export limits have pushed Chinese developers to build capable models with fewer resources. He estimates some Chinese open-weight models are six to 12 months behind models from leading U.S. companies, including OpenAI and Anthropic. Lee says Chinese companies mainly aim to make money, and that market competition, rather than a coordinated government plan, has driven their move toward open models. He also says Chinese companies tend to use resources carefully, while U.S. companies spend more to attract top researchers and provide computing power.
Companies still struggle to put AI to work
Many companies have not yet figured out how to use artificial intelligence effectively, despite its promise to boost productivity. AI can process large amounts of data quickly, but it can make mistakes and miss the nuances of human communication. Companies also struggle to fit AI into their workflows in ways that support human workers rather than replace them. As a result, some businesses have seen productivity gains while others are still working out how to use the technology.
AlphaPai brings its investment research AI platform to global markets
AlphaPai announced its international launch at Hong Kong FinTech Week × StartmeupHK 2026 on November 13, 2026. The company says its China-market product serves more than 120,000 professional users across over 8,000 institutions and asset managers. Its PaiWork research workstation and PaiMeeting assistant help investors gather information, review meetings and produce research. Tools include skills for analyzing companies, preparing management meeting questions and reviewing changes in earnings estimates, with coverage of Asian markets.
Study finds ultrasound AI overcounts lung B-lines
A study at Skåne University Hospital in Sweden found that an automated artificial intelligence tool counted more pulmonary B-lines than expert physicians. Researchers assessed the Venue GO point-of-care ultrasound system from General Electric Healthcare in 55 hospitalized patients with acute respiratory symptoms, mainly acute heart failure. The study collected 672 clips using a standard 12-zone scanning method, and 574 clips were suitable for statistical analysis. The researchers reported poor agreement between the software and the physicians, largely because the tool overestimated B-line counts.
Shield AI opens Bengaluru office to adapt Hivemind for India
Shield AI said Bengaluru is its second office in India, after New Delhi, and that it plans to hire engineers in autonomy, flight software, simulation, systems integration, and testing. India managing director Sarjan Shah said the new office will give the company an engineering team to build for local customers. The move follows an October test in Nevada where two combat drones used Hivemind to conduct coordinated maneuvers beyond visual range. JSW Defence is building a V-BAT production facility in Hyderabad as part of a $90 million investment, and India selected V-BATs and Hivemind licenses for the Indian Army in January.
Sources
- Baldwin Group CEO: The biggest AI disruption in insurance is the advisor
- Executive Perspectives: Iowa leaders share their AI outlooks
- The pathetic tyranny of artificial intelligence
- United Airlines won’t let AI agents book your flight. Expert says you shouldn't let a bot do that for you, anyway
- Opinion | America’s AI bubble is a Ponzi scheme
- ‘Chinese companies are born to make money’: AI pioneer Kai-Fu Lee on the 3 biggest myths about China’s AI race efforts
- Companies still don't quite know what to do with AI
- AlphaPai Launches its AI-Native Investment Research Platform for Global Markets at Hong Kong FinTech Week
- Point-of-Care Ultrasound Artificial Intelligence Overcounts Pulmonary B-Lines Compared to Clinical Experts
- Shield bets Bengaluru engineers can tune Hivemind autonomy software for India
Comments
Please log in to post a comment.