#243: After A Nobel Prize, What’s Next for CRISPR Cas9, & More
1. After A Nobel Prize, What’s Next for CRISPR Cas9?
This week Drs. Emmanuelle Charpentier and Jennifer Doudna made history, winning the Nobel Prize in Chemistry for the development of a gene editing tool, CRISPR Cas9. Many academic institutions are experimenting with CRISPR Cas9 and have published thousands of research articles expanding on its possibilities.
Currently, most clinical trials are ex vivo, the editing taking place outside of the body. We believe in vivo gene editing is the next frontier.
This week Intellia Therapeutics (NTLA) presented data at the Oligonucleotide Therapeutics Society (OTS) showing the successful suppression of TTR, the gene that causes a buildup of abnormal proteins called amyloids, in non-human primates. Human trials should start by the end of this year.
While the technology is nascent, we are inspired by the possibility that gene editing will deliver one-shot cures for both rare and chronic diseases.
2. Autonomous Taxis Are Launching This Year

In several weeks, Waymo plans to make history by launching the first commercial autonomous taxi service. After many years of trial and error, Waymo will let the general public ride in its fully driverless cars with no safety operators. Waymo One, currently serving a select group of consumers willing to sign non-disclosure agreements, is operating in fully driverless mode only 5-10% of the time. Now, Waymo is ready to roll out a fully driverless service.
While many automakers, including Audi, Nissan, BMW, GM, Toyota, and Volvo, initially aimed to launch autonomous cars this year, none has succeeded, leading some to wonder if autonomous cars are decades away. Waymo is proving that autonomous driving is possible now. Having pushed its own launch back from 2019, Waymo’s management must be highly confident now about the performance of its technology, which once struggled to make left turns.
Waymo’s announcement also indicates the limits of its approach to autonomous driving. It plans to expand its commercial service outside of the current 50 square mile area in Arizona but will have to include safety operators once again. Waymo relies on LiDAR sensors and highly detailed maps, limiting the pace at which it can scale. In contrast, Tesla does not rely on LiDAR and highly detailed maps and, at launch, plans to roll out its autonomous service nationwide. Likely launching later than Waymo, Tesla plans to serve a much larger geographic region, collecting many more miles of real-world driving data than Waymo, a boon in the artificial intelligence race for the first nationwide autonomous taxi network. Following Waymo’s announcement, Elon Musk tweeted this morning that a limited beta version of Tesla’s full self driving features will be released on Tuesday next week.
That said, Waymo is kicking off a monumental transformation in mobility. ARK’s research suggests that, at scale, an autonomous taxi could cost just 25 cents per mile, less than half the cost of driving a personal car and one tenth the cost of a taxi today. According to our research, the present value of the next ten years of cash flow from global autonomous taxi systems is roughly $2 trillion.
Safer, cheaper, and more convenient, autonomous transportation is no longer a dream.
3. Does Daimler's EV Strategy Update Need Updating?

This week Daimler presented its electric strategy through 2030. In ARK’s view, the presentation highlighted that Daimler is far behind the competition and on the wrong trajectory. While Daimler will release its first EV platform for large vehicles in 2021 – eight years after Tesla launched the Model S – it will not have its own operating system until 2024. Further, Daimler’s commitment to EVs is not as strong as its marketing suggests because it is hedging its bet with hybrids: its “uncompromising” electric-first platform leaves room for an internal combustion engine and a two-speed gear box. ARK’s research suggests that traditional automakers should transition to battery electric vehicles as fast as possible, avoiding the distractions associated with hybrid technologies, if they are going to have a chance at success.
4. Square Just Bought $50 Million of Bitcoin

Bitcoin is beginning to impact corporate balance sheets. Following Microstategy’s announcement a few weeks ago, Square announcedthat it will add bitcoin to diversify its liquid assets, the direct investment “[providing] a way to participate in a global monetary system, which aligns with the company’s purpose.” The $50 million investment approximates 1% of Square’s total assets and 2.5% of its cash and equivalents as of the second quarter.
Perhaps more important, Square has open sourced the processes and procedures enabling its bitcoin purchase. In Square, Inc. Bitcoin Investment Whitepaper, it documented the trading, custody, insurance, and accounting practices associated with the purchase. As CEO Jack Dorsey noted on Twitter, Square shared the documentation to encourage and enable other companies to join the movement.
5. China is Moving Closer to the Full Roll-Out of Digital RMB
On Friday, China announced that it will give away 10 million digital RMB in 50,000 digital “red packets” containing 200 yuan (~US$30) to Shenzhen residents. Based on the SCMP report, any resident in the Luohu district of Shenzhen can register for the lottery through iShenzhen, a blockchain-based public services app. This Sunday, the winners will receive access to the official Digital RMB app and will be able to spend the 200 yuan at 3,389 designated shops in the Luohu district.
During a speech at the Sibos banking and financial conference earlier this week, Deputy Governor Fan Yifei said that the People Bank of China (PBOC) already had processed 1.1 billion yuan (US$162 million) of digital RMB covering 3.13 million transactions in pilot programs. In these programs, the PBOC granted more than 113,000 personal digital wallets and 8,000 corporate digital wallets, according to SCMP.
In the absence of clear guidelines for the official launch of digital RMB, Fan has shared a few points of consideration in the past. First, broader distribution of digital RMB is likely to go through a two-tier system, leveraging both commercial banks and third-party payment platforms to prevent unintended increases in the money supply. Second, as a substitute for physical cash, or M0, digital RMB will not earn interest as the PBOC aims to prevent disintermediation from the existing banking system.
As digital RMB displace physical RMB, China will be able to monitor economic activities in real time, track suspicious transactions, and deliver “helicopter money”, if necessary, to end users. Like any other sovereign-backed digital currency, however, digital RMB should not be considered in the same realm as decentralized currencies like bitcoin. See our past writings on bitcoin and cryptocurrency here to learn the differences.
6. Is AMD + Xilinx a Good Marriage?
Hot on the heels of Nvidia’s announcement to acquire ARM, AMD is looking to buy Xilinx, the world’s leading designer of FPGAs (Field Programmable Gate Arrays). Unlike specialized chips, FPGAs can be reprogrammed, making them popular in industries such as telecom, automotive, and internet of things.
AMD likely wants to combine forces with Xilinx to turbocharge its penetration of the data center. While AMD’s Epyc processors have captured more than 10% of data center CPUs, neither AMD nor Xilinx has made any meaningful inroads into accelerators. Nvidia’s GPUs dominate the accelerator market with more than $4 billion in annual sales.
We believe the problem with this deal is that neither AMD nor Xilinx has strong traction in data center acceleration. AMD’s Radeon group has invested little in deep learning. Xilinx’s AI software stack is more mature thanks to acquisitions like DeePhi, but its Versal chip is too new and exotic for customers to take seriously. Xilinx boasts of 14,000 developers, paling in comparison to Nvidia’s 2 million.
Neither AMD nor Xilinx is large enough to be the center of gravity in the data center today. We doubt that the combined entity will be large enough.