Weekly Dose of Optimism #110

SSI, Neo, Solar Saturation, Replit Nuclear Clock

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

If you study the capital flows and announcements of the world’s largest asset managers, you’ll notice that private credit has become one of the most popular asset classes of 2024.

With Percent, accredited investors can participate in this growing asset class and tap attributes like:

• High yield potential (up to 20% APY) plus cash flow from monthly interest payments

• The option to redeploy capital frequently with shorter-term investments

• A wide swath of deals, ranging from small business loans, merchant cash advances, trade finance, consumer loans,  and more across the US, Canada, and Latin America

With the market near all-time highs, many investors turn to private credit because its returns are historically less correlated to public markets, which can stabilize a portfolio during a downturn. Percent is the easiest way to start investing in private credit, with minimum investments of $500 and average deal terms of just 9 months. 

Explore Private Credit and grab a $500 bonus on your first investment.

---

Hi friends 👋,

Happy Friday and welcome back to our 110th Weekly Dose of Optimism.

Dan here — it’s been a busy week. Aside from writing the Weekly Dose, my day job is running a company called Create. I started Create because I think way too few people take creatine — something like 2-3% of the population. I think that number should be much closer to 100% than 0%. We’re starting to make a dent. We now have well over 100K customers and have introduced over 50K folks to creatine for the first time.

This week, we announced a $5M Series A led by Unilever Ventures. The goal is triple down on what’s working, launch new creatine-forward products, build out a small team, and expand deeper into retail. The Information covered the round here.

I want to say thank you for reading Not Boring. Writing the Weekly Dose allowed me to not take a salary from Create for a year, re-invest that capital back into growth, and frankly stay sane as I dealt with the ins-and-outs of building a CPG business. As a thank you, you can get 50% off your next order using optimism50 — now until the end of the weekend (code auto-applied at checkout when using link): get 50% off here.

Let’s get to it.

---

(1) Exclusive: OpenAI co-founder Sutskever's new safety-focused AI startup SSI raises $1 billion

Kenrick Cai, Krystal Hu and Anna Tong for Reuters

Safe Superintelligence (SSI), newly co-founded by OpenAI's former chief scientist Ilya Sutskever, has raised $1 billion in cash to help develop safe artificial intelligence systems that far surpass human capabilities, company executives told Reuters.

There’s that old story about Picasso. Picasso is sitting in a café and a woman recognizes him and requests a drawing. Picasso obliged, quickly creating a small drawing, and then handed it to her. As he did, he said, "That will be $10,000." The woman, surprised, asked, "But it only took you a few minutes to draw this!"

Picasso replied, "No, it took me 40 years."

That’s kind of what is happened at Safe Superintelligence (SSI). The company has no product, a relatively vague roadmap, and just a small team of 10 yet it was able to raise $1B at a reported $5B valuation. But the team didn’t just raise the round overnight — co-founders Ilya Sutskever, Daniel Gross, and Daniel Levy have been working towards this for years. Specifically with Ilya at Google and OpenAI and Gross as the founder of Cue (acquired by Apple) and one of the most active investors of this AI boom.

Safe Superintelligence is the company’s name, mission, and entire product roadmap. The company is designed to make a straight-shot at safe superintelligence, without short-term commercial or product cycle pressures. I guess we’ll have to just wait and see how that strategy plays out.

(2) They Built This Robot For Your Home | 1X Technologies

From S3

1X Technologies is making humanoid robots for the home. Its latest robot, Neo, is apparently ready for mass production and pilots within in the home, and will even be available for sale as early as next year.

Neo is unique from other humanoid robots in two major ways. First, Neo is designed for the home, meaning that it can handle a multitude of different tasks and environments within the home. Second, it’s designed for everyday safe collisions (bumps, swipes, brushes, etc). Humans are constantly safely colliding with and within our environments, but heavy, rigid robots have not traditionally been designed for such collisions. That’s is an issue within the home.

1X has been in this space for over a decade now, and introduced an earlier humanoid robot EVE back in 2017. EVE comes with a wheeled base and is already pretty widely used in commercial settings like factories and warehouses.

We’ll see how Neo holds up in the market. Elon has claimed that Optimus will eventually cost $20K and tackle many of the same household use cases, while Unitree’s G1 already costs just $16K. No matter who wins, we continue to be amazed that the question on robots seems to be: “who will win and how cheaply can they do it?” versus “will these things ever work?”

(3) Can Solar Costs Keep Shrinking?

Tomas Pueyo for Uncharted Territories

If we want to know whether solar electricity costs will continue shrinking in the future, we need to break down its costs and figure out to what extent each one can shrink.

We’ve all been the beneficiaries of one of the most impressive cost curves in history. Since 1975, the price of solar photovoltaic (PV) panels has been shrinking around 12% per year. That decrease in price is, in part, the result of a massive increase in solar capacity (ie installs.) Simply put, “the more solar panels we produce, the cheaper they get.”

The question is: can we keep it up or have solar costs plateaued? According to Pueyo, there is still some room for cost improvements. First, there is no reason to believe that we’ll stop installing more solar panels, and in doing so, the cost of solar panels will continue to decline. Second, technological improvements like thinner or more efficient panels should continue to drive down costs. And finally, the whole ecosystem around solar should become more efficient — everything from the cost of capital to permitting costs — as solar becomes more ubiquitous and an even more important component of our energy mix.

Let’s keep driving down that curve until we have solar energy too cheap to meter.

(3a) Solar will get too cheap to connect to the power grid

From Ben James

The cost of solar panels is plummeting, and this will flood the power grid with cheap electricity. But that’s just Act 1. We won’t stop building solar at the limits of the grid - we’ll build a lot more.

OK, one more bullish solar think piece to share — this time from Ben James. James’ essay takes Pueyo’s argument a few steps further: not only will solar continue to get cheaper and cheaper, but it’ll be so cheap and abundant that it’ll exceed grid capacity, and we’ll have to use off-grid storage and intermittent processes.

James and Pueyo agree that solar costs will continue to plummet. Panels are getting cheaper, which will lead to more installs, which will lead to cheaper panels, and so on and so forth. The cost curve will keep marching down and to the right, while the capacity curve keeps driving up. James argues that this will eventually lead to grid saturation…grids will be flooded with cheap solar, to the point it doesn’t even make economical sense to buildout more grid connectors.

Ultimately, this will lead to more off-grid storage and use of solar. Excess solar will be stored in local batteries, avoiding transmission grid bottlenecks. And this cheap, off-grid solar can power processes like fuel production and industrial heating that can operate intermittently.

If you are to believe Pueyo and James, the future looks as bright as the sun.

(4) ‘Nuclear clock’ breakthrough paves the way for super-precise timekeeping

Elizabeth Gibney for Nature

Physicists have demonstrated all the ingredients of a nuclear clock — a device that keeps time by measuring tiny energy shifts inside an atomic nucleus. Such clocks could lead to vast improvements in precision measurements, as well as new insights into fundamental physics.

Tick-tock on the nuclear clock, but the party don't stop.

A nuclear clock is a timekeeping device that measures time using energy shifts within an atomic nucleus. Recently, physicists have taken a significant step toward developing a nuclear clock by measuring the frequency of energy shifts in thorium-229 nuclei with unprecedented accuracy, 100,000 times greater than previous efforts.

The team used a laser frequency comb to measure the energy shifts in thorium-229 nuclei embedded in a crystal, synchronizing these shifts with the world's most accurate atomic clock. Developing a nuclear clock is important not only because it allows for ultra-precise timekeeping, but also because it could enhance measurements in fundamental physics and help detect phenomena like dark matter.

(Yes, I did put on Tik Tok by Kesha while writing this. Oh-ooh, oh-ooh, whoa-oh)

(5) Announcing Replit Agent

Packy here. I gotta admit, with all the love that the AI-native code editor Cursor has been getting since it announced its $60M raise, I’ve been getting a little jealous of people who know how to code. Cursor is getting rave reviews from developers, but its main use case is making actual developers better and faster, not letting non-coders like me build things. To use it to make apps and websites, you need to set up a lot of other stuff in addition to writing code, none of which I know how to do.

So I was very happy to see our portfolio company, Replit, announce Replit Agent. The new product lets you build websites and apps simply through prompting. It imports the right libraries, sets up databases, and deploys for you, in addition to writing the code.

I tried it out to build a website for Not Boring, and while I need to tweak it to get the design right and actually import the content I want in there, it was able to build me something functional in 10-15 minutes, including a content management system that takes substack links to import my essays and put them on the blog page. I’m not going to win any engineering or design awards, but it’s wild to be able to get something from my brain, into software, and onto the internet within minutes.

The gap between idea and execution continues to shrink. What a time to be alive.

---

Have a great weekend y’all.

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

Thanks for reading,

Packy + Dan