The finest line of poetry ever uttered in the history of this whole damn country was said by Canada Bill Jones in 1853, in Baton Rouge, while he was being robbed blind in a crooked game of faro.
George Devol, who was, like Canada Bill, not a man who was averse to fleecing the odd sucker, drew Bill aside and asked him if he couldn't see that the game was crooked.
And Canada Bill sighed, and shrugged his shoulders, and said, 'I know. But it's the only game in town.' And he went back to the game.
― Neil Gaiman, American Gods
As mentioned briefly in edition #62, I re-listened to Peter Attia’s interview of Rick Johnson (I highly recommend it, there’s so many big ideas and actionable advice in it), and it provoked this thought:
Listening to scientists talk about the details of their field has a double effect on me.
It’s both a reminder of how complex it all is and how much we don't know, but also of how much we do know and the incredible amount of detail we know it in. We often know what is going on mechanistically atom-by-atom.
These two things combine synergistically and 2+2 = 5 (or whatever McKinsey lingo means that it’s a good pairing).
It's common for the uninitiated to show skepticism about a field because they know that first part (“it’s all so complex and difficult”), but they're missing that second part (“here’s five mountains of evidence on this topic”) to balance out and calibrate that skepticism.
So the skepticism usually overshoots and they end up skeptical of the wrong things, or to the wrong degree, throwing out the rosy-cheeks little baby with the bathwater.
⨳ On the Internet, everybody is one degree of Kevin Bacon away.
At first, you read this and think it’s a joke.
Then you read it again, and think it’s an opportunity.
But finally, you see that it’s also dangerous.
☑︎ Hello to Jeremy S., who has “read all 62 editions on the day they’re released”. That's real dedication (or masochism?). True OG!
APIs All the Way Down (there’s even a podcast version if you want to listen rather than read — Packy’s so thoughtful)
Some highlights and thoughts:
There’s an old Picasso fable that goes something like this:
A woman approaches Picasso in a restaurant, asks him to sketch something on a napkin for her, and tells him that she’d be happy to pay whatever it’s worth. He obliges, scribbles something quickly, and asks for $10,000. “$10,000!?” the woman replies in shock, “But you just did that in 30 seconds!” “No,” Picasso tells the woman, “It has taken me 40 years to do that.”
That’s one way to think about an API. APIs let companies leverage years of other companies’ work in seconds.
Is that not good writing? Right?
Hiring has traditionally been one of the most important things a company does. Picking the right API vendor is like the hiring decision on steroids. When a company chooses to plug in a third-party API, it’s essentially deciding to hire that entire company to handle a whole function within its business. Imagine copying in some code and getting the Collison brothers to run your Finance team.
Isn’t that explained so well?
I’m developing a writing crush on Packy. He and Byrne Hobart just have an incredible ratio of gem paragraphs…
And you know, when you call someone a good writer, you’re basically saying they’re a good thinker (not all good thinkers are good writers, but all good writers are good thinkers), and I don’t want to leave that implicit here.
Back to APIs:
What’s incredible to me about this graphic is that each logo represents something that a company would have had to build on its own previously, that it can now do by writing a few lines of code, and do better than they would have ever been able to in-house. [...]
This is the beauty of API-first companies. They allow customers to focus on the one or two things that differentiate their businesses, while plugging in best-in-class solutions everywhere else. [...]
Using a bunch of APIs that are really flexible, and figuring out good ways to connect them, leads to a combinatorial explosion of potential workflows. [...]
If the number of potential connections between APIs increases exponentially as more are added, companies have a near-infinite ability to create unique chain-link systems of coherent actions out of the existing primitives. Instead of a linear value chain in which using commoditized components in each step limits the number of places left to differentiate and create value, the Request/Response Model lets companies differentiate
I also recommend this one. I listened to the podcast version on a walk, and it’s a good overview of some of the historical background and research on psychedelics, and looks at the company Compass Pathways, who’s trying to commercialize Psilocybin therapies:
For more on psychedelics research, MAPS.org is probably a good place to start (you can donate to support their research/advocacy).
‘Amazon Becomes World’s Largest Corporate Purchaser of Renewable Energy’
I remember ˜10 years ago when Amazon was attacked by Greenpeace and other environmental NGOs because it was lagging behind other tech giants when it came to renewable energy. Always nice to see things move this quickly in the right direction:
Amazon today announced 26 new utility-scale wind and solar energy projects totaling 3.4 gigawatts (GW) of electricity production capacity, bringing its total investment in renewable energy in 2020 to 35 projects and more than 4 GW of capacity — the largest corporate investment in renewable energy in a single year. These new projects will make Amazon the largest-ever corporate purchaser of renewable energy.
Amazon has now invested in 6.5 GW of wind and solar projects that will enable the company to supply its operations with more than 18 million megawatt hours (MWh) of renewable energy annually. This is enough to power 1.7 million U.S. homes for one year. These projects will supply renewable energy for Amazon’s corporate offices, fulfillment centers, and Amazon Web Services (AWS) data centers that support millions of customers globally. (Source)
They’ve set a goal of reaching net-zero carbon emissions across all their businesses by 2040. A sub-goal was powering all Amazon infrastructure with 100% renewable energy by 2030, but they’re now on a trajectory to get there by 2025 instead. That’s the kind of guidance that more companies should over-deliver on…
DoorDash CEO Pulls an Oprah
DoorDash just IPO’ed, and as is the norm lately, its shares went up 80%+ on the first day. Tony Xu, the CEO, pulled a nice mensch move for the occasion:
I'm excited that our very first investment from today's milestone as a public company will go to the Merchants and Dashers on our platform. [...]
Today, our most active Dashers will find a cash bonus in their account, from $300 up to $20,000 (in their local currency). Also as part of our Main Street Strong campaign, Dashers will be eligible for additional rewards in 2021. More details will be shared in the coming weeks. (Source)
Speaking of DoorDash (and let me throw in Airbnb, while at it — note who the emails below are going to and the year):
Semiconductors in Europe: 13 EU Countries Join Forces to do… something
Europe’s share of the 440-billion-euro ($533 billion) global semiconductor market is around 10%, with the EU currently relying on chips made abroad. [...]
Security concerns regarding some foreign governments have also added to worries about relying on foreign chips used in cars, medical equipment, mobile phones and networks, as well as for environmental monitoring. [...]
The group will reach out to companies to form industrial alliances for research and investment into designing and making processors and look into funding for such projects. [...]
The signatories include Belgium, Croatia, Estonia, Finland, Greece, Italy, Malta, the Netherlands, Portugal and Slovenia. (Source)
I’ll be curious to see what comes of that. I’m certainly not rooting against it, as a more diversified global semiconductor industry would lead to a more stable and resilient world. And more investments in R&D can produce breakthroughs that we can all benefit from and build on top of…
I just hope it won’t all be top-down and deals with existing big companies, and that they’ll also work on setting the right incentives for startups.
SpaceX Starship next to Space Shuttle (no banana for scale)
A recent test of the huge Starship concluded with an impressive explosion, but that’s beta-testing for you. Things go wrong, but you still learn a lot, and you could have 99% of things go right but everybody will remember that 1%-massive-explosion… ¯\_(ツ)_/¯
Full video of the launch (it’s long, but if you just click around on the timeline you’ll easily find the good parts)
Anyway, Steve Jurvetson, an early investor and board member of Tesla and SpaceX, posted this nice composite giving a better sense of scale of this monster (because against a blue background in the sky, it could be any size):
I get what they were going for with the bare metal, the whole essentialism thing, but I still think it would look better with a coat of paint. Maybe it’s just me ¯\_(ツ)_/¯
DNA → RNA → Ribosome → Protein
I was thinking of how to explain the mRNA vaccines to my parents (they asked), and think this image does a good job.
We’ve basically found a way to package mRNA and inject it into our bodies in a way that makes it to cells. In cells, the ribosome reads that mRNA and produces the proteins that we’ve encoded.
What we’ve encoded is the spike protein that sites on top of the SARS-CoV-2 virus. As our immune system comes into contact with that spike protein, it learns how to recognize it and make antibodies that bind to it.
That’s my understanding of the 101 of it, anyway. Image source.
I wonder if we’ve ever had another medical breakthrough that went from being theoretical to being deployed to billions of people and likely saving millions of lives in such a short amount of time:
Up until December 2020, no mRNA vaccine, drug, or technology platform, had ever been approved for use in humans, and before 2020, mRNA was only considered a theoretical or experimental candidate for use in humans.
As of December 2020, there were two novel mRNA vaccines awaiting emergency use authorization as COVID-19 vaccines (having completed the required eight-week period post final human trials) – mRNA-1273 from Moderna, and BNT162b2 from a BioNTech/Pfizer partnership. (Source)
Very impressive.
While on this topic, Zeynep Tufekci wrote an interesting piece about both how impressive these mRNA vaccines are (back in April even top experts like Dr. Fauci expected we’d need 18-24 months, and the FDA was ready to approve vaccines with 50% efficiency… “we have is a vaccine ready at half the time the most optimistic timeline projected, with twice the efficacy hoped for. It’s remarkable.”
The question she raises is about how effective a single dose of the two-dose Pfizer vaccine is, and whether we should consider single-dosing certain groups since it looks like there won’t be enough doses for everyone for a while:
Eyeballing (we don’t have the underlying data), the first shot seems to have 65-80% efficacy in the full group—but we don’t know for how long that lasts. Hence, we have durability of immunity data—for three more months, to day 119—only with the booster shot.
Given the shortages, the potential for single dosing—at least for some populations—or more time between the doses is an important question. It’s possible that a single dose—one that can cover twice the number of people—would provide a significant benefit to the recipients, though we would be unsure about whether the immunity protections last as strongly three months later. [...]
Is 60-80% efficacy vaccination for 200 million is better than 95% for 100 million? Given how terrible the next few months look to be, how do we prioritize what we do first? Can we space the booster a bit while we wait for longer-term data?
While on the topic of biology (I know, my segues are really smooth):
Just to mention, human genome is 3.3 Gb in size. HIV virus is only 9.7 kb. Largest known virus genome is 2.47 Mb (pandoravirus salinus). Largest known vertebrate genome is 130 Gb (marbled lungfish). Largest known plant genome is 150 Gb (paris japonica) (Source)
That’s quite a range, from the Porcine circovirus type 1, which is only 1,759 base pairs and is the “Smallest viruses replicating autonomously in eukaryotic cells”, to that 150,000,000,000-base-pair monster Paris Japonica that doesn’t look anything out of the ordinary when you see it.
