• One more positive step toward artificial general intelligence…
  • A new pivot for Meta and its metaverse…
  • How to short bitcoin…

Dear Reader,

I hope you enjoyed our small detour in The Bleeding Edge last week, as we explored some different investment strategies in order to deal with the crazy markets that we find ourselves in these days.

I enjoyed writing about these different topics, and I appreciate the latitude to explore some different ideas with you. The world will return to growth and advancement, but for now, we’ll have to do our best to survive the current turmoil.

Speaking of turmoil, the proposed deal between Twitter and Elon Musk continues to deliver excitement and entertainment.

This has been a front-row seat for us to view a real-time mergers and acquisitions (M&A) negotiation, with a backdrop of the last two years of social media manipulation, censorship, and debate about freedom of speech.

On Friday, Musk notified Twitter that he’ll be terminating the acquisition due to the material breach of certain provisions of the agreement.

More specifically, Twitter has failed to provide Musk with the requested information concerning:

  • Information related to Twitter’s process for auditing the inclusion of spam and fake accounts in monetizable daily active users (mDAUs)

  • Information related to Twitter’s process for identifying and suspending spam and fake accounts

  • Daily measures of mDAU for the past eight quarters

  • Board materials related to Twitter’s mDAU calculations

  • Materials related to Twitter’s financial condition

As we review the above requests for information, these are perfectly logical things to ask for. Not only are they important for understanding the overall business, but this information also provides the most critical inputs to determine an accurate valuation for Twitter.

As a reminder, Twitter claims that fake or spam accounts represent less than 5% of all accounts on Twitter. This is what Musk’s original valuation was built upon.

Yet a joint audit conducted by two firms back in May estimated that more than 19% of all active Twitter accounts are fake or spam accounts. That is a very material difference compared to “less than 5%.”

It is completely normal for any buyer of a company that finds evidence of such a material difference in business metrics – and is rebuffed in their efforts to receive full disclosure from the target company – to want to terminate the deal.

Yet that’s exactly what the financial media have proclaimed. “Musk is pulling out!”… “The deal’s off.”…

And several have gone so far as to suggest that this was all a clever ruse by Musk to sell off a chunk of his Tesla shares so he could take some money off the table.

Yes, the deal may come to an end, but Musk still wants Twitter. And this is still a negotiation. Twitter refused to play ball and share material information about its business that was required under the terms of the deal.

And there’s a reason Twitter hasn’t done so… the numbers aren’t good and would reveal that either Twitter was materially withholding information about its business from its shareholders, or was willfully turning a blind eye to what has been happening on the social media platform.

Either way, it is a failure of both the executive management and the board of Twitter… And Musk knows it.

Further, he is using this to his advantage. The letter of termination is just another salvo in the negotiations. Twitter refused to disclose; Musk moves to terminate. In reality, the negotiations will continue between the two parties.

Twitter will insist the deal moves forward, but I don’t believe it has any chance that it would prevail in receiving the $1 billion breakup fee from Musk. It hasn’t met its obligations.

The “market” has already spoken as well. Twitter’s share price is down 13% since last Thursday and is now trading around $34 a share. That compares to the offer price of $54.20 a share. That implies that the market also believes that Twitter has a problem with fake accounts.

The reality is that Musk holds all the cards. Twitter is in a very weak place right now.

I don’t believe that it will prevail with the breakup fee, and I don’t see it faring well on its own if the deal falls through now that we have an idea of how bad the fake accounts are on the platform.

And that means that Musk has Twitter at the table to renegotiate at a price that reflects the value of the company based on the actual and mDAUs on the platform. If Twitter accepts, Musk wins. If Twitter declines, Musk still wins and can take his capital and invest it elsewhere.

That’s a great position to be in.

Yet another major breakthrough toward artificial general intelligence…

We have started to see developments around artificial intelligence (AI) accelerate in recent months. And despite the economic chaos, the field of AI has been on a tear this year.

Just last month, we talked about how advanced Google’s LaMDA and Meta’s new AI have become. In fact, there was talk about LaMDA possibly even being sentient.

This month, it’s OpenAI’s turn.

OpenAI just pioneered a clever approach to training AIs. This time, the challenge at hand was for an AI to play the massive online game Minecraft.

The significance here is that Minecraft is very open-ended. People can do pretty much anything they want. They can explore the world… Build structures… Hunt… Cook… Create new instruments and tools… the possibilities are nearly endless.

So there’s really no structure or logical path to gameplay in Minecraft. And that makes the game an incredibly complex environment for an AI to operate in.

But the team at OpenAI had an idea.

First, they created a group of neural networks – These are AIs modeled after the human brain. And they decided to “teach” these AIs how to play by watching videos on YouTube.

OpenAI took 2,000 hours’ worth of video and labeled it with text to explain what was happening and what the objects presented were. Then the OpenAI team fed the neural networks an additional 68,000 hours of Minecraft videos from YouTube. These were simply videos of people playing Minecraft.

With the first 2,000 hours as a foundation, the AI ingested the additional 68,000 hours and labeled the remaining hours of the video itself. Then it used that newly labeled content to learn how to play the game.

Once complete, the AI had a full grasp of how to interact and compete within Minecraft. It learned to become fully functional within the open-ended virtual world.

This alone is a fantastic breakthrough. And what’s even more exciting are the implications…

If we think about it, there are millions of hours’ worth of video on YouTube demonstrating how to do anything and everything we can think of. It turns out these are extraordinary resources that can be used to teach an AI how to do virtually any task.

Then I think about the rapid advancement of robotics, and it gets even more exciting… especially when I think about bipedal robots like Tesla’s Optimus.

Using the technique just demonstrated by OpenAI, we could train bipedal robots to do just about anything. Washing dishes… doing laundry… vacuuming the carpet – we could train a robot on simple videos of these tasks and the end result would be a fully functional home butler.

So there are some immensely practical applications for this.

And bigger picture, this is yet another major breakthrough toward artificial general intelligence (AGI).

As a reminder, AGI is when an AI is indistinguishable from a human and is empowered with “superhuman” capabilities. It can think, understand, learn, and apply its knowledge to solve any problem – just as we humans do – but in a fraction of the time and on any subject matter at all.

And not only does this development bode well for applications in the real world, but it also has implications in a metaverse. Most don’t realize it, but Minecraft is basically a metaverse.

If AI can be developed in a way to become functional in Minecraft, the same can be done for any other metaverse.

And this raises some questions. As Ais, or intelligent “bots,” interact inside of a metaverse, it will soon become nearly impossible to tell whether we are interacting with another human participant or with an AI.

Taking things one step further, these intelligent bots can actually become “real” economic actors inside of a metaverse, producing economic value for whoever created them. Imagine a world where we have intelligent avatars of ourselves working in a real or virtual environment, and producing something of value every day…

Given the advancements that we’ve seen year to date, we’re in for even more excitement before the year’s end.

Meta’s metaverse plans are taking shape…

Meta (formerly Facebook) just released more on the company’s vision for its metaverse. And it reveals an interesting pivot…

Longtime readers may remember when Facebook was actively pushing Libra back in 2019.

Libra was to be a stablecoin backed by a basket of fiat currencies. And Facebook’s master plan was that it intended Libra to serve as a global reserve asset.

Of course, Facebook ran into a mountain of regulatory trouble over Libra. And it ultimately sold all of the digital currency assets to Silvergate Capital earlier this year.

That said, Meta just announced a new digital wallet technology for use in its metaverse. And the wallet won’t just be for digital currency. It will also hold metaverse objects in the form of non-fungible tokens (NFTs), and will be used for digital identity and authentication as well.

So Meta is moving away from the idea of a global reserve stablecoin, but the company still very much intends to enable payments in a digital currency. It will just be for its metaverse, and will be somewhat of a walled garden, which is much less threatening to the regulators.

In addition, Meta announced a new online store called Meta Avatars Store. This is where people can buy clothing and accessories for their digital avatars for use within the metaverse.

I can imagine an environment where many will spend most of their working hours in a metaverse. Less money will be spent on clothing for going out in the real world, and there will actually be a massive market for clothing and fashion for our avatars to be worn in metaverses. I know it sounds crazy, but it is already happening, as we’ve seen in The Bleeding Edge.

I’m excited to see Meta’s approach to digital currencies in its metaverse. Will it use just one… Or will it allow people to use a small number of cryptocurrencies in its virtual world? Will it allow any purchased NFTs used in other metaverses… Or will it strictly limit them for use in its own?

I’ll continue to follow this story closely in the pages of Unchained Profits. There, we’ll watch for the most promising developments with investment potential

The nuances around the new Bitcoin ETF…

The U.S. Securities and Exchange Commission (SEC) just approved a new Bitcoin exchange-traded fund (ETF)… But it’s not the one we have been waiting for.

The industry has been working hard to get a Bitcoin spot ETF approved. In fact, it has been trying to launch a spot ETF for years.

Having a security like this would make it simple for all investors to gain direct exposure to bitcoin, through their normal equity brokerage accounts, without having to set up a digital wallet or have an account with a digital asset exchange.

A spot ETF would differ from the current Bitcoin ETFs in that they would actually hold the underlying asset.

However, the SEC has repeatedly rejected Bitcoin spot ETF applications. Instead, the regulators have only approved “futures” ETFs for bitcoin. These funds don’t actually hold bitcoin. Instead, they are constructed of derivatives designed to mimic bitcoin’s price movements.

So the new Bitcoin ETF that just launched is a futures ETF. It’s called the ProShares Bitcoin Short Strategy ETF (BITI). And as the name implies, this ETF allows investors to get downside exposure to bitcoin.

BITI allows normal investors to short bitcoin in the simplest way possible. When the price of bitcoin falls, BITI goes up. So to short bitcoin, all investors need to do is buy BITI.

While I would prefer to see the SEC approve Bitcoin spot ETFs, I am glad to see this security gain SEC approval. It’s a good thing for the market. And it is a way for all investors to be able to buy a single security if they want to gain exposure to bitcoin to the downside.

But I do want to offer a fair warning here: There’s a lot of expense involved with these futures ETFs. That eats into profitability significantly.

In other words, when Bitcoin falls in price, we shouldn’t expect BITI to go up by the exact same percentage. That’s because fees will weigh down its returns a little bit.

For this reason, investors have to time the market well to profit from instruments like BITI. These are not buy-and-hold investments.

For BITI to increase in value, Bitcoin would have to drop significantly and consistently over a period of days. This is the kind of security that can be useful as a hedge in the event of uncertain market conditions – and it should most likely only be held for a matter of days or weeks, not months.

Still, I see this as a positive development for the market. And I hope we’ll see a Bitcoin spot ETF follow.


Jeff Brown
Editor, The Bleeding Edge