ARK • Disrupt

During its Connect¹ event earlier this fall, Meta showcased new smart glasses that fuse technologies from MetaAI with Ray-Ban sunglasses to create a personalized AI assistant. In a compelling demo, Meta showed how users can look at an object, ask questions, and receive comprehensive answers instantly.

After using the Ray-Ban smart glasses for a week, we are deeply impressed. Powered by Meta’s large language model, Llama 2, the glasses respond to voice messages in an intuitive and nuanced way far superior to Siri and Alexa Voice.

Meta's AI assistant also seems to offer the first hands-free computing experience that sets it apart from alternatives. Without touchscreens, the Ray-Ban Meta AI sunglasses enable users to interact with their surroundings, access information, and complete tasks seamlessly. While a wholesale shift to voice-centric interaction has yet to be proven, Meta's integration of AI with smart glasses could be the first killer feature for smart glasses—and the most compelling case for hands-free computing yet.

In their respective earnings calls last week, Ford and GM announced delays in their electric vehicle (EV) plans, both explaining that the production of affordable and profitable EVs is proving more challenging than anticipated.² Are they suggesting that EVs are not ready for prime time?

Perhaps the confusion stems from the common assumption that their experience and scale would enable traditional automakers to dominate the EV market and that, by circular inference, their failure suggests that the timing is not right for EVs. Yet, while well-endowed traditional automakers are struggling, innovative companies like BYD and Tesla are scaling the production of affordable EVs profitably, undercutting legacy assumptions, and leading the way.

Pure-play EV companies seem to be disrupting traditional automakers. In our view, disruption is where the opportunity and action are, suggesting that delays in plans for aggressive EV investments could exacerbate the challenges to their future that traditional automakers are facing.

This week, ARK’s Multiomics Research team assessed the potential economic value of curing all inherited rare diseases, a goal associated with the Chan Zuckerberg Initiative’s mission to prevent, cure, and manage all rare diseases by the year 2100. To estimate the economic potential, we scraped data from OrphaNet³, a well-known database of rare diseases, using a combination of annotations provided by medical professionals and by AI LLMs (large language models) like OpenAI's GPT-4. Based on our data scrape, we developed a model that accounts for the following variables:

• Disease prevalence
• Disease severity
• Disease effect on life expectancy
• Disease mode of inheritance
• Estimated annual direct costs for each disease

Based on an estimated cost-per-quality-adjusted-life-year of $100,000, our model suggests that curing all individuals with inheritable rare diseases today would create ~$17.5 trillion in value. Our current model excludes avoided costs, like those associated with chronic care. Including them would increase the value created significantly. Despite the challenging path ahead, we believe advances in and convergences among artificial intelligence, gene editing, RNA silencing, cell therapies, and advanced small-molecule modalities will transform Chan Zuckerberg’s goal into a reality much sooner than even they seem to be anticipating. Stay tuned for our blog on this important topic!