The Fed, DNA Sequencing, More AI Developments

Don’t Fight The Fed

It’s an important time to practice investment patience. September brings another CPI release (Sept 13) followed by a Fed rate decision (Sept 21). The Fed has said many times they will continue to tighten until the inflation rate comes down. For now, we can get an early peek at the inflation numbers that will arrive at the Fed by looking at commodity markets.

Notably, international shipping costs 10x’d during COVID. Since then, they’ve fallen to just 4x the pre-COVID levels with a y/y reduction of roughly -50%. This matters because the cost of international shipping in our global economy eventually feeds into all prices. Global shipping rates are one of many leading indicators of inflation. Gasoline is almost back to where it started the year, down nearly 50% from the panic levels. Food, housing and oil market prices, unfortunately, remain elevated. For one housing data point, the median U.S. home price as of March 2022 represents a 16.3% y/y increase. As for August 2022, that y/y increase fell to just 6% y/y increase. If the median home price falls another 5%, the y/y growth rate will turn negative.

Until prices stabilize or fall, the Fed’s tightening activities make it a tough environment for most investors. The next rate hike will bring interest rates to the highest level since the ‘08/’09 crisis, which many people feel will be unsustainably high. In our view, the early answer to what happens next lies in the commodity markets. As of today, the commodity data presents a mixed bag of good and bad news. Hence, it’s a good time to practice investment patience.

DTCC + Distributed Ledger Technology

DTCC, the Depository Trust & Clearing Corporation, is one of those important backbones of the modern world that most people never hear about. DTCC currently processes around $2.4 quadrillion in global securities transactions annually. DTCC has a new crypto-inspired project called Project Ion that has recreated a portion of their settlement process on DLT (distributed ledger technologies) developed by Corda. Project Ion has a white paper that you can read here. There isn’t a token or crypto asset related to Project Ion, but the project uses crypto-inspired technology.

One thing I’ve said before about crypto that may end up coming true: crypto’s main function may not be to create crypto itself, like Bitcoin, but more to put pressure on existing structures to force them to improve through the threat of emergent competition. In other words, Bitcoin or Ethereum do not need to overtake traditional finance, AND traditional finance can adopt the best-of-the-best crypto technology to improve the current systems. DTCC’s ultimate goal includes paving the path for T+0 settlement, or immediate trade settlements. The current system allows for T+2, or two days to settle each trade.

DNA sequencing unlocks new opportunities

The cost to sequence the human genome was roughly $1mm in 2008. That cost fell to $100k in 2009, $10k by 2011 and $1k by 2015. In May 2022, Ultima Genomics raised $600m on the claim they can sequence a human genome within 20 hours at suitable quality standards for just $100. In public markets, Illumina is considered one of the industry leaders of the sequencing space. Regardless of who is providing the sequencing, this pricing structure unlocks a whole new world of opportunities to collect data that was previously inaccessible. This is a perfect example of how our world continues to innovate at an exponential rate. Once we can collect massive databases of our core code (DNA), we can link together via cloud computing to share the data and potentially use AI to train models to discover new core factors that influence our health from a root level.

For another example, “shotgun metagenomic sequencing” is a method of sequencing used to evaluate bacterial diversity and abundance from a single sample. This method is being used to analyze gut biomes which play a crucial role in health. We are at the data collection phase and have barely scratched the surface when it comes to analyzing what foods and behaviors impact our gut biome or what DNA markers lead to particular food or lifestyle sensitivities. All of this points to the potential to confidently target the root causes of diseases so that they can be addressed and monitored as early as possible.

Rapid AI Evolution

Regular readers know that I follow AI-generated art tools closely. When Dall-E was released by OpenAI, the company explicitly stated that they own the rights to the images. Since then, nearly a dozen competitors have created their own models, including Stable Diffusion, which is fully open source and requires that all images enter the public domain. In a matter of months, AI-generated art tools transitioned from exclusive to inclusive.

Since it is open source and very high quality, Stable Diffusion is quickly making its way into producing AI-generated videos, video game elements, fashion prototyping, and 3D-generated VR spaces via Unity. You can search other people’s prompts via Lexica.art. You can also buy successful prompts to better control your desired outcome.

As far as my understanding goes, AI is not the one doing the work. So when a company says “powered by AI”, they are usually overstating AI’s role. New revolutionary AI techniques, like the transformer model developed by Google in 2017, are what help create better models. Those models are essentially the “AI” that people reference. The computing power to run those models is considerable. It takes even more compute to train the models.

AI-trained models are being used to process natural language to service the restaurant and hospitality world. Panera bread uses voice AI it in its drive-throughs and Danny Meyer’s ConverseNow aims to automate the restaurant order-taking process, removing humans from that conversational process entirely. Some argue that AI tools like Stable Diffusion will smother human creativity, but I’d argue the opposite will happen. Open source tools like these will enable anyone to participate in the creative process by dramatically reducing the friction of turning ideation into creation.