• Fully self-driving cars will be on the road before year’s end…
  • Here’s a clue about the future Mars colony…
  • Our weather predictions are about to get more accurate…

Dear Reader,

There was an unbelievable breakthrough in securities regulations in the U.S. yesterday. And it will greatly benefit normal retail investors.

It is one of the last things I expected to write about on Election Day, particularly because this issue has been largely partisan…

The U.S. Securities and Exchange Commission (SEC) voted to amend the current regulations surrounding crowdfunding. More specifically, the SEC increased the amount that can be raised by private companies from nonaccredited investors:

  • Tier 1 Regulation A deals can raise up to $20 million

  • Tier 2 Regulation A deals will now be able to raise up to $75 million (up from $50 million)

  • Regulation Crowdfunding (Reg CF) deals will now be able to raise up to $5 million (up from $1.07 million)

And it is the last bullet point that has me the most excited.

Regulation A deals under the previous structure rarely ever raised $20 million or $50 million. Therefore, the raise to a $75 million cap won’t materially change much. And Regulation A deals are almost as much work as going public, which means a very limited number of companies pursue that route for a capital raise.

But Reg CF deals are different. They are designed for early stage private companies and are much easier than a Reg A deal. They are less expensive and employ a simpler process of filing to raise capital.

Reg CF deals were originally envisioned in the Jobs Act of 2012 as a way for small private companies to raise capital from nonaccredited investors. But the maximum raise in any 12-month period was limited to $1.07 million.

This low cap on Reg CF deals deeply and negatively affected high-quality early stage companies raising capital via crowdfunding. The amount was too small to have a meaningful impact on funding the business. $1.07 million typically isn’t enough capital to fund a technology or biotechnology company for even a year.

In short, the effort simply wasn’t worth it for most companies. And generally speaking, the best early stage companies went a more traditional venture capital (VC) route because of this dynamic.

But a $5 million cap changes the equation entirely.

Under good management, a $5 million pre-seed or seed round raise can fund a business for a good two or three years. That is enough time for a company to find the right product market fit, develop a go-to-market strategy, and potentially even reach free cash flow breakeven.

Once accomplished, the company then has some great options. It can return to a more traditional VC raise for its next round of funding and do so from a position of strength. Or it can choose to grow more conservatively alone while using its free cash flow to fund the business.

Another great alternative is to take in strategic growth capital from non-VC channels.

And the best part is that normal retail investors can participate in Reg CF deals. With a $5 million cap, thousands of nonaccredited investors can participate in a single deal.

As a simple example, if a group of investors each put in $500, 10,000 investors could participate in a Reg CF deal.

This is a fantastic development for investors, and it’s one I have long been a strong proponent of.

Accredited and nonaccredited investors should be able to invest their hard-earned money as they see fit. And over a multiyear period, I know of no other asset class that has more upside potential than early stage private investments.

No matter what happens over the next few days, at least we have this victory for investors.

This one simple change has the potential to dramatically change the structure of how early stage capital formation takes place. It can actually shift the power toward entrepreneurs, innovators, and normal investors and away from the VC firms that take their pound of flesh with each round of capital raised.

I’ve been working hard for years to bring early stage investments to my subscribers, and this latest development with the SEC is going to dramatically improve my ability to do so.

We have a lot to look forward to in 2021.

Now let’s turn to our insights…

The rapid advancement of Tesla’s fully self-driving beta software…

We talked about the beta release of Tesla’s fully self-driving software two weeks ago.

To bring new readers up to speed, Tesla did a scrap and rebuild of its autonomous-driving artificial intelligence (AI) for this beta release. Tesla built its new AI from scratch, basing it on neural networks.

But the AI did not start from scratch on the data. On the contrary, Tesla used its database of billions of miles of self-driving data collected over years from its “fleet” of AutoPilot-enabled vehicles. The result was a newly architected, fully self-driving software package.

This software was pushed out to Tesla’s most cautious and committed drivers. These drivers have been testing the software on the road in a live beta over the past two weeks.

And the progress has been remarkable.

Tesla drivers have demonstrated their cars navigating complex and busy parking lots. They have reported the cars driving themselves at night and in the rain – conditions that have posed challenges in the past.

Tesla drivers have even shown that their cars can drive themselves on gravel roads. That’s especially challenging for the AI because there are no road markings to use as reference points.

It’s not perfect yet. Tesla CEO Elon Musk made that clear right up front. There have been instances where the driver needed to intervene and grab the wheel.

But the brilliant part of this beta is that Tesla is taking in the driving data and feedback from its own passionate customers to improve the software in real time.

In fact, the team is pushing out software upgrades every few days, sometimes hours. One beta tester said that the software is improving “shockingly fast.”

And that speaks to Tesla’s innovative model.

Unlike other carmakers, Tesla’s cars are designed around the computer architecture. They are software-oriented. And they can be upgraded as simply as we update the operating systems on our computers, phones, and tablets. It’s a remarkable system for innovating and iterating quickly.

And what makes Tesla’s approach unique in the industry is that its software is being designed to function in every environment. Tesla is not preprogramming geographic areas like other companies are doing.

This is evident when we look at Waymo One, which we talked about a few weeks ago as well.

Waymo mapped out about a 50-square mile area in Phoenix, Arizona. Its cars are preprogrammed to drive themselves in this area because they know every street, stop sign, and intersection.

But that’s where it stops. Waymo’s cars can’t drive outside of this geofenced area without a safety driver present.

Tesla isn’t geofencing anything. Its goal is the full monty – full autonomy door to door.

Tesla is putting this software out in the wild, leveraging its existing user base to accelerate the learning process, collect more data, and put changes to work quickly. It’s incredibly exciting to watch.

And I am very encouraged by these early results. I think Tesla is on track for a complete implementation of its fully self-driving software by year’s end. It may not be 100% by that time, but it will be pretty darn close.

Musk’s vision for Mars…

Talking about way out there…

I recently enjoyed Elon Musk’s presentation at the 2020 Virtual Mars Society Convention. In it, Musk shared his vision for the Red Planet.

Musk’s plan is to create a self-sustaining city on Mars. He envisions a city that could survive even if supplies from Earth stopped coming in. Of course, that would require the city to be able to produce everything it needed on-planet.

And here’s where it gets really interesting…

Musk’s vision is for a self-governing colony on Mars. He envisions a society that’s not tied to any nation-state on Earth. In fact, Musk said that this colony would not recognize international law.

Obviously, this has major philosophical and political implications. There will surely be an issue for some nation-states on Earth that would not be comfortable with “ceding” power or influence over its citizens even when they are off-planet (i.e., not on Earth).

But Musk is serious about this. Much of the work SpaceX is doing is geared toward making such a self-governing society possible.

And cleverly, Musk has already been laying the groundwork. A perfect example is SpaceX’s recent Starlink satellite internet service. Inside of Starlink’s terms and conditions, buried in the legalese, is this statement:

For Services provided on Mars, or in transit to Mars via Starship or other colonization spacecraft, the parties recognize Mars as a free planet and that no Earth-based government has authority or sovereignty over Martian activities. Accordingly, Disputes will be settled through self-governing principles, established in good faith, at the time of Martian settlement.

How’s that for forward-looking? This shows us that Musk plans to launch a Starlink satellite constellation around Mars as well – or at least above the inhabited areas of Mars.

And remember, SpaceX has already announced plans to send its first Starship to Mars by 2024. This will likely be a mission to send out supplies in advance of a future manned mission to the planet.

Assuming that SpaceX hits its target, 2024 is going to be an exciting year for space exploration. After all, that’s the same year the U.S. plans to launch its return mission to the Moon. It is moving fast.

Many people have passed Musk’s ideas off as crazy. But there’s no doubt in my mind that he is a visionary and is capable of making it a reality.

And I absolutely expect to see many of his ambitions regarding Mars come to pass. We have the technology, the know-how, and the company (SpaceX) to make a manned mission to Mars possible within the next decade.

I can’t tell you how excited I am about this. This reinvigoration of NASA and the aerospace industry is so inspiring and motivating.

It will hopefully inspire the next generation of engineers and scientists. And it will also have the added benefit of giving human civilization a safety net if things go horribly wrong here on Earth – man-made problems or otherwise.

Yet another breakthrough in AI…

This last topic may sound arcane, but please bear with me. The implications are widespread.

An incredible research paper just came out of Caltech a few days ago. A team of researchers applied AI to partial differential equations (PDEs).

These mathematical equations are widely used to calculate the flow of liquids and air. Naturally, this is necessary when it comes to designing aircraft, spacecraft, and even boats. And we also use PDEs to help make projections involving motion.

This is an area that I used to be very familiar with. I worked with PDEs when I studied aeronautical and astronautical engineering at Purdue University.

The thing is, PDEs are notoriously complex and hard to solve. In fact, solving PDEs typically requires supercomputers.

And there is a limit. The more complex the calculation, the more difficult it is for even a supercomputer to solve. That’s the problem Caltech set out to tackle with AI.

By applying a form of AI called deep learning, the team at Caltech demonstrated that AI could solve PDEs 30% more accurately and up to 1,000 times faster than the traditional methods. Incredible.

And it can be done without the need for massive supercomputers.

This is going to impact many different industries.

Now that AI can tackle PDEs quickly and efficiently, we’ll see major advances in things like seismology, materials science, and even weather forecasting.

One of the most difficult problems that we have not been able to solve is actually modeling the Earth’s atmosphere and weather patterns. We don’t have enough computing power on Earth to do it accurately. But the breakthrough at Caltech has the ability to change that.

As consumers, we won’t be able to see, touch, or feel these advances. After all, this breakthrough doesn’t come in the form of a 5G-enabled smartphone.

But we will reap the benefits of breakthroughs in numerous industries as a result of this new AI.

And for what it’s worth, this will definitely go a long way toward improving our weather forecasts.


Jeff Brown
Editor, The Bleeding Edge

Like what you’re reading? Send your thoughts to [email protected].