Hi friends 👋,

Happy Friday and welcome back to our 117th Weekly Dose of Optimism. Another week, another full slate of big stories. Arizona is making chips, Arkansas is busting at the seams with lithium, Berkley researchers are developing new carbon capture materials, AI launch velocity picks up, Stripe is doing some splashy crypto M&A, and blindness is being…cured? You just love to see it ;)

Let's get to it.

Today’s Not Boring is brought to you by… Speakeasy

As the old saying goes, “Get someone to use your product, get a user for a day; get someone to integrate your API, get a customer for life.” Or something like that. 

I’ve been a fan of the API-first business model since writing about it back in 2020. Half the battle is building APIs that handle mission critical but non-core work for your customers. The other half is making it as easy for them to integrate your APIs as possible. Speakeasy can help integrate API users 50% faster. 

Creating a frictionless API experience for your partners and customers no longer requires an army of engineers. Speakeasy’s platform makes crafting type-safe, idiomatic client libraries for enterprise APIs easy. That means you can unlock API revenue while keeping your team focused on what matters most: shipping new products. Make SDK generation part of your API’s CI/CD and distribute libraries that users love at a fraction of the cost of maintaining them in-house. Let Speakeasy handle your mission-critical, non-core work so you can do the same for your customers. 

If you know what any of those terms meant, and if those pain points sound familiar, you should check out Speakeasy.

Try Speakeasy for Free

(1) TSMC’s Arizona Chip Production Yields Surpass Taiwan’s in Win for US Push

Mackenzie Hawkins for Bloomberg

Writing the Weekly Dose is hard. We have the whole thing written and ready to go, and then boom: Bloomberg reports that TSMC’s Arizona manufacturing facility is producing chips at 4% higher yields than similar facilities in Taiwan.

That is huge. We’re optimists, and frankly, we were a little worried about TSMC’s ability to manufacture here. Less skilled labor, safety and management issues, all the things that normally get in the way of building things. TSMC pushed back its plans to start full production here by a year, then pushed back the launch date for the second fab.

And then they go and totally redeem themselves. Now, these are 4nm chips, not the 3nm chips TSMC makes at home, but a 4% improvement in yields shows that we can manufacture these complex little wafers here, might unlock billions in additional CHIPS Act funding, and might push TSMC to open all six possible fabs in Phoenix.

The way things are heading, we’re going to produce oodles of chips powered by gigawatts of nuclear energy while eagles fly overhead, kids eat apple pie, and someone terraforms the Arizona dessert in the background. Out with the longshoremen, in with the onshoremen.

(2) Evaluation of the lithium resource in the Smackover Formation brines of southern Arkansas using machine learning

Knierim et al in Science

We calculated that there are 5.1 to 19 million tons of lithium in Smackover Formation brines in southern Arkansas.

About 6 weeks ago, we covered the lithium boom in the the Smackover Formation region of the United States. That story covered how one particular entrepreneur was planning to use Direct Lithium Extraction, a methodology he developed back in the 90s, across one particular region of the southeast referred to as the Smackover Foundation.

A paper recently published in Science confirmed that that entrepreneur is on the right track. The researchers identified a substantial lithium resource in southern Arkansas. Using an AI model trained on geological, geochemical, and temperature data, the team predicted lithium concentrations across the region and estimated the formation contains between 5.1 and 19 million tons of lithium, or about 35% to 136% of the current U.S. lithium resource estimate. Lithium boom baby!

If the U.S. is going to usher in a clean, renewable future, it’s going to need access to a lot more lithium — the essential mineral for building EV batteries and other clean tech. And boy does it look like Arkansas has us covered. God blessed America.

(3) Carbon dioxide capture from open air using covalent organic frameworks

Zhou et al in Nature

Here a porous, crystalline covalent organic framework (COF) with olefin linkages has been synthesized, structurally characterized and post-synthetically modified by the covalent attachment of amine initiators for producing polyamines within the pores. This COF (termed COF-999) can capture CO₂ from open air. COF-999 has a capacity of 0.96 mmol under dry conditions and 2.05 mmol g^–1 under 50% relative humidity, both from 400 ppm CO₂. This COF was tested for more than 100 adsorption–desorption cycles in the open air of Berkeley, California, and found to fully retain its performance.

Having access to abundant lithium would likely reduce our CO2 emissions in the future. But we’ll still need technologies that capture and remove the CO2 that has and will continue to make its way out into the world.

Luckily, Cal Berkeley researchers have developed COF-999, a new porous, crystalline organic material that efficiently captures carbon dioxide directly from open air. The material functions like a specialized sponge made of a porous material, with tiny holes throughout. These pores are designed to selectively attract and trap carbon dioxide molecules from the air as it passes through. Then, when the material is gently heated, it releases the captured CO₂, allowing COF-999 to be reused for continuous carbon capture.

Someone like Terraform Industries or Valar Atomics can turn the CO₂ into fuel, that fuel will be used to power planes, and those planes can keep sprinkling COF-999. Bad time to be carbon dioxide. Good time to be a human.

(4) Science Announces Positive Preliminary Results For Vision Restoration In Pivotal Clinical Trial

Science Corporation, a leader in brain-computer interface (“BCI”) technology, has announced the preliminary clinical trials results for its PRIMA retinal implant. The patients in the trials had lost their central visual field, which makes them unable to read and struggle to recognize faces. The results showed that the PRIMA implant restored real form vision in these patients such that sequences of letters can be read with a clinically meaningful improvement of letter acuity.

John 9:25 - Whether he is a sinner or not, I don't know. One thing I do know: I was blind but now I see.

Science (the Corporation, not the field of study) released very promising preliminary results from clinical trials of its PRIMA retinal implant. The PRIMA system is a wireless, autonomous implant that works with special glasses and a pocket processor to enhance visual clarity. Essentially the glasses capture an image and that image is transported to an implant in the patients eye that allows them a to “see” a version of that image.

The implant restored meaningful vision to individuals who had lost their central visual field, enabling them to read sequences of letters and even longer texts. After 12 months, patients showed an average improvement of 23 letters (4.6 lines) in visual acuity, with the best patient improving by 59 letters (11.8 lines). Meaning, the trial participants went from being blind to being able to read short passages.

Brain-computer interfaces (BCI) are going to get a lot better in the coming years. One day, I’m sure we’ll all have some chip planted in our brains that makes us superhuman. But it’s awesome to see that initial beneficiaries of these BCI advancements are the people who need it most, be it quadriplegics or the visually impaired.

TL;DR Science and technology are giving sight back to the blind.

(5) Two Big AI Launches

From Runway: Introducing Act-One

I’ve never really understood how animation works. In my head it always goes back to flipbook animation. That’s the technique where sequential drawings on pages are flipped rapidly to create the illusion of motion. You likely made a flipbook animation in first grade. It was exhausting and shitty. So when I see animated films today, I always imagine a team of artists flipbooking their way through an intricate 2 hour long movie. Reader, that’s now how animation works today and it hasn’t worked this way in a long, long time. But if you’re like me and that’s your base-level understanding of animation than this new product from Runway is going to blow your mind.

Runway’s Act-One generates expressive character animations using only an actor's video and voice—no extra equipment required. By capturing the nuances of facial expressions and performances, Act-One translates real-life acting into digital characters, even when proportions differ from the source.

It’s always kinda funny in the animated movies when the character kind of resembles the actor (although it’s even funnier when the character is totally different than the actor.) Like Maui in Moana. The Rock was born to play Maui and Maui was likely created to be perfect for The Rock. This tech takes that connection to a whole other level and does so in a low-cost, no equipment way.

From Anthropic: Computer Use

Anthropic introduced a new capability called "computer use" for its Claude 3.5 Sonnet AI model. Computer Use allows Claude to interact with computers in ways similar to human user. It can perform tasks like moving the cursor, clicking buttons, typing text, and switching between windows or applications. You know, things we mere mortals regularly do on computers. Claude is essentially interpreting screenshots to understand what's on the screen and to determine how to complete given tasks.

You don’t have to extrapolate the demo of computer use too many years down the line to see how a technology like this could be disruptive. The majority of knowledge work done today consists of interpreting information on a screen and then moving our cursor or making a few keystrokes based on that that information. AI models have been good at the interpretation part for a couple of years now, but computer use bridges the gap between interpretation and action in a real-world helpful way like no model has before. It’s a little clunky today and for most people the prompts needed to complete a simple task would take longer than actually just doing the task, but you can pretty easily imagine the future when much of the prompting is abstracted away and descendants of computer use are automating much of the knowledge economy’s work.

Fine by me: “Hey computer use, please write and publish this week’s edition of the Weekly Dose of Optimism.”

EDIT: After writing the above passage, Anthropic added another feature to Claude: the ability to write and run code. Claude taking down another important segment of the knowledge workforce. Told you it was hard to keep up with everything.

Bonus: Stripe Confirms Plans to Acquire Stablecoin Platform Bridge

Emily Mason for Bloomberg

Stripe Inc. co-founder and Chief Executive Officer Patrick Collison confirmed in a X post on Monday that the payments company plans to buy stablecoin startup Bridge…Stripe will pay $1.1 billion, according to a person familiar with the matter who asked not to be identified discussing private information.

Packy here. On Monday, Bridge confirmed the rumors that Stripe would acquire the company for $1.1 billion. This is both the largest acquisition in Stripe’s history and the largest acquisition in crypto history.

Bridge is both a Not Boring Capital portfolio company and founded by Duke grads, so personally, I’m psyched about the deal, but more importantly, this is a really big deal for the way that money moves around the world. Stablecoins are just better, faster, cheaper infrastructure for moving money, and Stripe has both the scale and talent to make stablecoins the default much more quickly.

As Stripe co-founder Patrick Collison tweeted, “Stablecoins are room-temperature superconductors for financial services. Thanks to stablecoins, businesses around the world will benefit from significant speed, coverage, and cost improvements in the coming years.”

Something like this was inevitable. When I wrote about Stripe back in August 2020, I predicted that it would want to eventually rebuild economic infrastructure from the ground up, and acquiring Bridge is a big step in that direction.

Now, we’re going to get to see how much more creative leverage entrepreneurs can unlock if they get ~2% of their revenue back to spend on more productive uses.

Have a great weekend y’all.

Thanks to Speakeasy for sponsoring! We’ll be back in your inbox on Tuesday.

Thanks for reading,

Packy + Dan