This is NOT investment advice. Investing in startups is risky whether it is via StartEngine or any of the other platforms. Liquidity may or may not be an option for an extended period of time.
Let's see if we can toss out a few names on a monthly basis for people to debate and evaluate and see if this is useful. At times, we have only one name or zero. We understand some may feel the need to "do something" or "pull the trigger" or "just buy something to allocate and diversify" but we don't see it that way.
The types of firms that typically catch our attention are generally based on seeking a potential "bet a little, win a lot" approach where we seek potential "winner takes all or most" outcomes. We tend to focus on those and avoid battles with low barriers to entry UNLESS the firm is simply executing at a very high level and clearly advancing to a few different potential exits.
First, a couple of recent posts here on this platform:
When is CHECKMATE - Bitcoin Update
The Submerge - Ether Merge Debacle
Perhaps a little blurb about what we will not invest in could be helpful. When seeking a potential "winner takes all" and vice grip lock on certain economic value, we try to be very strict. We have to feel it and understand it. If we don't "get it" in terms of how and why the pieces fit together, then we pass. This can happen frequently with tech, biotech, medical, and retail. We aren't smart enough to know all of these details for each industry and products. If we can't connect all the dots we pass. If the business relies on hype and marketing we pass.
Let's sneak in an interesting update on Reg CF and the sky is falling from Pitchreviews.com with the image shown below:
Our take on the above is the following:
- this shows Reg CF funding rising at a much higher rate of change than that of the traditional VC funding, though should be noted that the "law of large numbers" is a factor here (going from $2M to $5M as an example versus going from $200M to $210M)
- nonetheless, regarding the first point, this still shows that Reg CF perhaps has more of a growth jolt to it than regular VC
- it is still very, very early here and we only have about 6 years of data on Reg CF here we are analyzing
- very interesting times here, so you can see the strong growth and can recall the status of the world in that 2018-2020/21 period versus the status of our world since then
- neither Reg CF nor VC funding are going away and in fact we expect both to increase and wrestle more control away from Wall Street and DC
- it is still very early for Reg CF
AtomBeam - We've commented extensively already about this opportunity. Our entry is 8M. A+ is ongoing and doing quite well already.
AtomBeam now has a commercial customer (Ericsson), a pilot with a large satellite player (ViaSat), and a pilot with a oil company underway. Let's first take a quick look at a list, provided by AtomBeam, of the products, IP, and where things stand:
A lot going on there. Here's a way possibly to make it simple as an investor. AtomBeam has three products. They're selling one now, one is very credibly in development, and a 3rd is very new, untested, but another revolutionary breakthrough potentially. Think of it in terms of a timeline and likelihood of achieving success and exit for your money. DoD contracts help with a foundation, Neurpac brings in commercial revenue in the short to intermediate time frame, and the other two products are all on top of all of this and TBD as to when they would produce revenue.
So - what the heck is Lightweight Codeword Model (LCM)? Well, we were taking some time to absorb it so we could answer that question ourselves ultimately.
Another slide below from an AtomBeam presentation:
Here's what we can process right now as we expect to learn more about LCM as time goes on:
- LCM can potentially reduce the burden in terms of time, energy, and power to run AI models
- LCM can help, possibly, make things possible such as "running AI stuff on your iPhone"
- LCM can prepare and package data so that it is easier to ingest and work with the data
- LCM can try AI stuff on the fly whereas the current models can only "retrain the AI" to "make it smarter"
The "stuff" that this thing called AI needs to eat . . . . . AtomBeam's new LCM might be able to cook it up real nice for the AI industry and help them train their AI agents on the fly.
How is AtomBeam any different than current "Compression" tech?
