In December of 2024, the Chinese start-up DeepSeek shocked the world with the release of an AI model that appeared much cheaper to make and run than those from its American rivals. The company also open sourced its AI, meaning it released the blueprints of its model to the public.

Our guest this week is Misha Laskin. His start-up Reflection AI looks to be the prime counterweight to DeepSeek and a host of other open source Chinese models. Laskin argues that open source models can be just as good as the models developed by the likes of OpenAI, Google and Anthropic and that the West needs this to be the case.

Reflection has raised $2 billion and is valued at $8 billion, although the figures in AI have become so lofty as to almost feel meaningless at this point. That said, the company was able to raise so much money because of the pedigree of its team with a number of engineers, like Laskin, coming from DeepMind.

I have, in full confession, not been paying enough attention to open source models and the ways in which the Chinese models seem to have become the basis for a lot of corporate work in the US. Laskin dug in deep on this topic, and hopefully you’ll feel more up to speed after listening to the episode.

You can find the Core Memory podcast on all major platforms and on our YouTube channel.

Enjoy!

Our show is sponsored by Brex, the intelligent finance platform. Like thousands of ambitious, innovative companies, we run on Brex so we can spend smarter and move faster. And you can too. Learn more at www.brex.com/corememory

The podcast is also made possible by E1 Ventures, which backs the most ambitious founders and start-ups.

This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.corememory.com/subscribe

Well, I wrote a book and made a movie about this week’s guest, so he must be fascinating. (Otherwise, I wasted six years of my life.)

Will Marshall has done things beyond supplying me with material. He’s the co-founder and CEO of Planet Labs. For those who don’t know, Planet changed the aerospace industry forever by lowering the cost of satellites and proving that they could be mass produced. It has surrounded the Earth with hundreds of satellites that photograph and analyze all our planet’s landmass every day.

Go on. Sign up. We bring you these people for free. Help us, help you.

In this episode, we get into Planet’s history, what these satellites do and why they’re so important.

Marshall is a deep thinker on many areas and has lived in an unconventional manner. So we also get into much, much more, including the communal house scene in the Bay Area, what Marshall sees as the three major shifts that will occur over the next decade and how he balances his idealism with being a capitalist.

In conclusion, Marshall is fascinating. I did not waste six years of my life. And When The Heavens Went on Sale makes for a great Christmas present.

Seriously. Let’s make the world smarter.

Our show is sponsored by Brex, the intelligent finance platform. Like thousands of ambitious, innovative companies, we run on Brex so we can spend smarter and move faster. And you can too. Learn more at www.brex.com/corememory

The podcast is also made possible by E1 Ventures, which backs the most ambitious founders and start-ups.

This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.corememory.com/subscribe

Over the past week or so, Elon Musk has started hyping up the launch of the long-, long-awaited Tesla Roadster 2.

Musk appeared on the Joe Rogan podcast last week and said he hoped to unveil the car before the end of the year. (When it goes on sale is another story.) He suggested the car might fly. (Okay?) And he said that it would almost certainly be the most memorable launch of any product in history. Which is a very Musk thing to say and - also in keeping with Musk - possibly true.

At the Tesla shareholder meeting today, Musk again promised to wow with the Roadster 2 unveiling. This promise came in response to a shareholder who asked if he could, in fact, have the first Roadster 2 VIN. “Well, I guess it’s according to whoever put down their deposit in that sequence,” Musk said.

Well, we are here with a special podcast to reveal exactly who stands to get the first Roadster 2 off the line – the investor Konstantin Othmer.

I ran into Othmer rather by coincidence this week, and he happened to be holding the receipts that show he wrote the first Roadster 2 check for $200,000 back in 2013.

This episode has some wonderful early Musk and Tesla tales plus the whole Roadster 2 backstory. Enjoy!

Our show is sponsored by Brex, the intelligent finance platform. Like thousands of ambitious, innovative companies, we run on Brex so we can spend smarter and move faster. And you can too. Learn more at www.brex.com/corememory

The podcast is also made possible by E1 Ventures, which backs the most ambitious founders and start-ups.

This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.corememory.com/subscribe

Joshua Steinman spent four years (2017-2021) working for President Trump and had a very broad remit. He shaped all cyber, telecommunications, cryptocurrency, and supply chain policy.

It’s fair to say the Washington press corps did not adore Steinman. He was often portrayed as a young, brash Silicon Valley-type who bubbled over with ambition and lacked the usual political decorum.

Despite how the press corps felt, Steinman did important work on several fronts. Before and during his time in Washington, he helped create deeper ties between the U.S. military and Silicon Valley in a bid to modernize the technology at the Defense Department’s disposal. He also sounded repeated alarms about how vulnerable the U.S. infrastructure is to cyber attacks, particularly those originating in China.

These days Steinman runs Galvanick, a company aimed at hardening the technology infrastructure of industrial companies and operations.

Steinman is opinionated and then some. In this episode, he will claim that Trump is among the smartest humans on Planet Earth. Some of you will be okay with this. Some of you will hate this. I look forward to your comments.

Beyond Trump, we get into Steinman’s unusual career, the state of U.S. military technology and security, the cyber Cold War between the U.S. and China and a host of other light topics. Enjoy!

The Core Memory podcast is available on all major platforms, including Apple, Spotify and YouTube.

Our show is sponsored by Brex, the intelligent finance platform. Like thousands of ambitious, innovative companies, we run on Brex so we can spend smarter and move faster. And you can too. Learn more at www.brex.com/corememory

The podcast is also made possible by E1 Ventures, which backs the most ambitious founders and start-ups.

This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.corememory.com/subscribe

Celine Halioua and her company Loyal are on track to deliver a drug next year that could help dogs live longer.

Loyal’s therapy is aimed at senior dogs (10+ years of age) that weigh more than 14 pounds. It’s a pill that the dogs will take daily and that’s designed to extend the dogs’ lifespan by at least a year. To get to this point, Loyal conducted a massive clinical trial with 1,300 dogs, and the FDA has liked what it’s seen so far.

Halioua joins the podcast this week to chat about her unique approach to cracking the longevity field.

Loyal has been betting that it will be easier to prove that longevity drugs work (and get regulators on board) by starting with dogs instead of humans. The company has been testing promising longevity compounds and now has three therapies in its drug pipeline aimed at our canine friends.

I’ve known Halioua for several years now. She’s one of the deepest, most pragmatic thinkers in the longevity field and approaches her work without the hype and false promises that often accompany some of our live forever gurus.

We get into her life, her work, some of the oddities of running a company in San Francisco and what it’s like to be in bio-tech during the great AI hype era.

Enjoy!

The Core Memory podcast is sponsored by Brex, the intelligent finance platform. Like thousands of ambitious, innovative companies, we run on Brex so we can spend smarter and move faster. And you can too. Learn more at www.brex.com/corememory

The podcast is also made possible by E1 Ventures, which backs the most ambitious founders and start-ups.

This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.corememory.com/subscribe

We know that there are trillions of dollars’ worth of minerals sitting on the ocean floor. The big question is whether humans should start hoovering them up.

Our guest this week is very much Pro Hoover. He’s Gerard Barron, the head of The Metals Company, which has spent years preparing to become a major player in the seabed mining industry. The company has found a spot in the Pacific Ocean that’s full of black, baseball-sized nodules rich in nickel, copper, cobalt, and manganese. In short order, The Metals Company plans to send vehicles down to gather the nodules up and then refine them.

We work hard to bring you these guests. Please subscribe and help support Core Memory. Thanks!

The pro case here is that the nodules are quite literally sitting atop the ocean floor. You don’t need to burrow into the seabed and wreak the usual environmental havoc associated with mining. You don’t need people laboring under dire conditions. And these nodules are so mineral rich that you don’t need the typical amounts of refining to get at the good stuff.

The con case is that we don’t know a ton about what goes on down there on the ocean floor in terms of animal life. Us humans could be triggering yet another environmental catastrophe on our way to harvesting what we desire.

Barron is well aware of the criticisms against seabed mining. John Oliver, among others, has gone hard at him and The Metals Company. And Greenpeace thinks Barron might be Satan.

Meanwhile, the U.S. very much wants to become a seabed mining power, as it attempts to blunt China’s dominance in critical minerals and rare earths. And, of course, the modern world depends on things like nickel, copper, cobalt, and manganese, and seabed mining looks like a very efficient way to get more of them.

We get into the pros and cons of this new field in gory detail on the pod. Some of you will be satisfied with Barron’s philosophy. Some of you won’t. In either case, you’ll come away better educated on the history of this industry and the technology driving it.

Enjoy!

The Core Memory podcast is sponsored by Brex, the intelligent finance platform. Like thousands of ambitious, innovative companies, we run on Brex so we can spend smarter and move faster. And you can too. Learn more at www.brex.com/corememory

The podcast is also made possible by E1 Ventures, which backs the most ambitious founders and start-ups.

This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.corememory.com/subscribe

Jed McCaleb grew up in Arkansas where he lived in a cabin in the woods that had no electricity or running water. Now, he’s a billionaire building a space station and funding some of the world’s most adventurous science.

America remains a thing, I guess.

McCaleb, 50, has been at the center of several major technology movements. Back in the peer-to-peer glory days, he released the eDonkey application for music and file-swapping and had the pleasure of being sued by the major record labels. Then, in 2010, he launched the Mt. Gox bitcoin exchange, which dominated the crypto world until it turned into a story of mega woe.

A couple of years later, McCaleb developed the Ripple protocol before moving on to build yet more crypto innovations. This work made McCaleb fabulously well-to-do. (He’s worth $3 billion . . . if you believe Forbes.)

McCaleb has used his fortune to back a number of start-ups and philanthropic endeavors. When Elon Musk pulled out of supporting OpenAI, McCaleb helped backfill the financial vacuum. He’s also been a major investor in Max Hodak’s Science Corp., which is developing brain computer interface technology and in the rocket maker Firefly Aerospace.

McCaleb’s real blockbuster investment is Vast Space. The start-up is building a commercial space station designed to be the successor to the International Space Station, which is very much on its last legs. McCaleb says he’s prepared to put $1 billion of his fortune into Vast to make sure it happens.

My conversation with McCaleb travels across these various tech eras. He comes off, at least to me, as an oddly down-to-Earth guy for someone who invests in such a variety of wild ideas.

Enjoy!

The Core Memory podcast is sponsored by Brex, the intelligent finance platform. Like thousands of ambitious, innovative companies, we run on Brex so we can spend smarter and move faster. And you can too. Learn more at www.brex.com/corememory

The podcast is also made possible by E1 Ventures, which backs the most ambitious founders and start-ups.

This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.corememory.com/subscribe

We have a proper treat this week. Core Memory special correspondent Eryney Marrogi comes on the pod.

Marrogi is a scientist and soon-to-be doctor who has followed the rise of the gene editing tools at humanity’s disposal. He wrote a piece for us a couple of months ago on “Baby KJ,” the child who received a customized, life-saving gene therapy in record time. Now we get into what Baby KJ means for the future of gene therapies.

The gene therapy field is advancing quickly but remains costly and technically complex. Marrogi breaks down how hopeful people should be – or not – that new techniques can be applied to more than super rare, single gene problems.

We also get into the ups and downs of the DIY Bio movement, Marrogi’s work creating gene-edited mosquitoes and the vaping crisis he sees with America’s youth, who are taking in unfathomable levels of nicotine.

Enjoy!

The Core Memory podcast is sponsored by Brex, the intelligent finance platform. Like thousands of ambitious, innovative companies, we run on Brex so we can spend smarter and move faster. And you can too. Learn more at www.brex.com/corememory

The podcast is also made possible by E1 Ventures, which backs the most ambitious founders and start-ups.

This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.corememory.com/subscribe

In the 1970s, scientists created the first wiring diagram of a worm brain. To do this, they sliced up a worm, imaged the slices under microscopes and then reconstructed a map of the worm’s neurons and their synaptic connections. Wonderful. (It was 302 neurons and really a map of the worm’s central nervous system.)

We haven’t come terribly far since then. It took until 2024 for scientists to create a full map – aka a connectome – of an adult fly brain (140,000 neurons) and researchers now dream of completing similar work for a mouse brain and one day perhaps a human brain with its 80 billion neurons and one trillion connections.

The belief is that if we have a proper wiring diagram of the brain, we will understand the brain and how it works much better.

Things have been slow and hard because mapping something like a mammalian brain requires a lot of laborious work, tons of computing power and some method for labeling all the neurons and their connections in a comprehensible and useful fashion.

E11 Bio was founded in 2021 to try and develop new techniques for mapping brains faster and cheaper. And today marks a big moment in the organization’s history. The E11 Bio team has published a paper – in conjunction with loads of high-profile contributors – detailing the success of its techniques. (There’s much more info on E11’s web site here and here.)

In very simplified terms, E11 can put viruses in the brain that carry proteins to neurons. Those proteins then distribute markers across the neurons that make them light up in different colors under a microscope. This technology has made it much easier to find individual neurons and trace their connections.

E11’s co-founder and CEO Andrew Payne was kind enough to come on the podcast to discuss the organization’s work and its brain mapping process. We also got into why this all matters, the challenges ahead and where neuroscience and artificial intelligence overlap.

Enjoy!

If you care about science and this type of in-depth reporting, please subscribe and support our work.

The Core Memory podcast is sponsored by Brex, the intelligent finance platform. Like thousands of ambitious, innovative companies, we run on Brex so we can spend smarter and move faster. And you can too. Learn more at www.brex.com/corememory

The podcast is also made possible by E1 Ventures, which backs the most ambitious founders and start-ups.

This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.corememory.com/subscribe

In this episode, Dan Wang comes on the show to discuss his new book Breakneck: China’s Quest to Engineer the Future.

I’ll start by noting that the book is fantastic, and you should read it. It’s a well-researched, vibrant account of how China became dominated by its engineering culture. The country has displayed an unmatched ability to build over the past forty years, and Wang traces the scale of these accomplishments in detail. He also documents how pervasive this engineering mindset is by diving into the one-child and zero-Covid policies, and the brutally efficient ways they were carried out.

Subscribe to the Core Memory podcast here and on all major podcast platforms.

Wang contrasts China’s engineering-first culture with the US’s regulation-first culture. China’s top politicians are mostly engineers. The US’s top politicians are mostly lawyers. Wang argues that the US once built like China until the 1960s came, and the US began regulating itself into a collective torpor.

As you’ll find in the book and hear in our discussion, Wang is not a China propagandist. Far from it. He offers a sober look at the pros and cons of both China and the US and points out that the two cultures have remarkable similarities.

In this episode, we explore the book and much beyond it, discussing what hope, if any, the US has of competing against China in the coming century.

Enjoy!

The Core Memory podcast is sponsored by Brex, the intelligent finance platform. Like thousands of ambitious, innovative companies, we run on Brex so we can spend smarter and move faster. And you can too. Learn more at www.brex.com/corememory

The podcast is also made possible by E1 Ventures, which backs the most ambitious founders and start-ups.

This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.corememory.com/subscribe