Skip links

Explore AI’s potential and pitfalls: Five Links for the Week

Explore AI’s potential and pitfalls: from enhancing healthcare, debunking robot butler myths, to navigating generative AI’s rocky reality in tech.

#1  When the AI party is over, it will come down to whether AI can create business value or not.  Here is Forbes with some simple suggestions about how this might happen.


The article from our wonderful friends at Forbes, authored by Aaron Reich and Jillian Moore, emphasizes the critical journey from experimenting with generative AI to realizing substantial business value. With the proliferation of AI technologies, organizations are eager to leverage AI for growth and efficiency. Yet, many are stuck in the early phases of AI readiness—Exploring and Planning—without significant ROI. The key to advancing lies not just in adopting AI for optimization but in innovating beyond automation, as exemplified by Klarna and Page Group, who have successfully harnessed AI to revolutionize customer service and data analysis, respectively.

The article highlights the importance of aligning AI with business goals, emphasizing that a successful transition requires a people-first approach. This involves equipping employees with new skills, fostering trust in AI technologies, and ensuring inclusive and equitable enablement. Trust, the cornerstone of value realization, must be nurtured through responsible AI use, robust data practices, and thoughtful governance. The piece underscores the need for comprehensive organizational governance to mitigate risks and foster innovation responsibly.

To accelerate AI value realization, the authors suggest three strategies: adopting people-first principles, focusing on desired outcomes rather than specific AI functionalities, and clearly defining the organization’s vision for AI, balancing innovation with practical value creation. Effective AI integration transcends technology, requiring organizational readiness, strategic foresight, and a commitment to ethical principles to ensure trust and sustainable value in the AI era.

#2  But questions remain.  Would you want an AI doctor who gives great answers to you questions and the best of the best in bed side manner but gets the diagnoses and treatment wrong much of the time

Harvard Health

A Harvard Health article explores a study comparing the empathy and quality of medical advice provided by AI, specifically ChatGPT, versus physicians. Initially, headlines touted AI’s superiority in these areas, stirring both interest and skepticism. The study involved evaluating responses to 195 patient questions from an online platform, comparing those generated by ChatGPT to those from doctors, with ChatGPT being rated significantly higher in both empathy and quality. ChatGPT’s responses were deemed better in quality 78% of the time and were found to be more empathetic than those of physicians by a wide margin. However, the study had limitations, including its failure to assess the accuracy of the answers or whether patients would accept advice from AI. It also raised concerns about the potential bias in the evaluation process, given the noticeable difference in response lengths between ChatGPT and physicians, possibly influencing perceived empathy. The article concludes that while AI could serve as a tool for enhancing medical consultations, its reliability and acceptance by patients require further investigation. It emphasizes the need for additional research to fully understand AI’s role in patient care.

#3. So what’s wrong with AIs? They are all liars


The excitement surrounding ChatGPT and similar generative AI technologies has given way to a phase of skepticism and realization of limitations, described as a “trough of disillusionment.” Initially hailed as revolutionary, generative AI has encountered significant hurdles, including errors in content creation, concerns over intellectual property rights, and environmental impacts. These challenges have dampened the early enthusiasm, causing some startups to struggle and reevaluate their strategies. Despite early success in areas like coding, the broader application of generative AI has proven complex, with companies finding it difficult to integrate these technologies reliably into customer-facing solutions. Ethical issues, such as bias and discrimination, remain unresolved, casting doubt on the practicality of these tools beyond novelty uses. However, the industry remains optimistic, viewing these setbacks as part of the natural progression of technological innovation. As generative AI continues to develop, there is confidence that with further refinement and ethical considerations, it will find its place in solving both mundane and complex tasks, signaling a future of sustained investment and advancement in AI capabilities.


#4.  But we are still publishing a Gasman’s Guide to Personal Robotics in a few months!


The dream of AI-powered robot butlers remains distant, with technological, economic, and practical challenges hindering their development. Despite significant advances in AI, translating these into physical robots capable of performing household tasks has proven difficult. High-profile demos, like the one by startup Figure showcasing a humanoid robot handling simple tasks, underscore the gap between concept and real-world application. These robots often follow narrow, scripted routines and lack the mobility and adaptability required for varied domestic chores.

Historical aspirations for robot butlers, inspired by early 20th-century fiction, confront reality’s complexities. Tasks easy for humans, like picking up objects or navigating homes, remain challenging for robots due to Moravec’s Paradox. This principle illustrates the disparity between robots’ abilities in digital tasks versus physical actions, where environmental unpredictability presents major obstacles. Robot arms and hands, crucial for tasks like cleaning or cooking, are still prone to clumsiness and limited by their inability to handle real-world variability with finesse.

Moreover, the economic feasibility of robot butlers is questionable. The cost of advanced robots remains prohibitively high for average consumers, with current models geared more towards industrial or commercial applications rather than household help. Despite optimistic market predictions, the reality is that robot butlers that can affordably and effectively take over domestic chores are still far from becoming a household staple. This situation leaves the vision of AI-powered domestic help more as a tantalizing prospect than an imminent reality.


#5. On motorbikes, Yamaha and AR


Yamaha is venturing into augmented reality (AR) for motorcyclists, building on their past work with HUD (head-up display) technology showcased as early as 2015. The concept of AR, which overlays digital information onto the real world, is gaining traction in various fields, thanks to advanced products like Apple’s Vision Pro and Meta’s Quest. Yamaha’s interest in AR helmets aligns with previous efforts by companies like BMW, which has been exploring AR in helmets and driving glasses for over two decades.

Yamaha’s latest patent moves away from the idea of fully screen-covered helmets to a more practical HUD approach. This design uses clear lenses that can display computer-generated images directly in the rider’s line of sight, addressing the dynamic nature of a rider’s vision affected by head movements and different riding postures. The system employs multiple tiny cameras aimed at the rider’s eyes, using infrared light to determine the gaze direction for accurately overlaying AR content.

Although the technology for AR helmets is advancing, Yamaha’s venture into this space raises questions about market readiness. Despite the technical progress, it remains uncertain if consumers are willing to invest in AR helmets, which are yet to hit the market. Yamaha’s efforts highlight both the potential and the challenges of integrating AR technology into motorcycle safety and navigation systems.

Leave a comment