The AI bubble of 2023
We are probably in the midst of an AI investment bubble. Public and private markets are both showing signs for concern.
AI VC money gets spent fast
Lots of celebrities and famous people also make money from things like venture capital and private equity and SPACs. If you are a VC or PE firm, there are two things you need to be able to do really well (among a few others):
1) Call investors to ask them to give you money, and
2) Call CEOs and ask if you give them money
Both of these are problems of getting other people to answer the phone when you call. That’s why famous people like Shaquille O’Neal and Kim Kardashian are also able to somewhat credibly add “investor” or “entrepreneur” to their chyrons on TV interviews; everyone is secretly a 12-year-old kid in the disguise of an adult and wants to hang out with models, actors, and professional athletes.
Elsewhere, investing trends come in hype cycles. It starts with the true wonks and inside-players who see something exciting well before it’s ready for primetime. Then it attracts smart money investors making early-stage bets. Then it gets to the massive VC funds who drop money from helicopters on every company they can find even remotely related to the trend. The last stage of the hype cycle is the celebrity investing & pumping, which is usually followed by SEC investigations.
Well, AI is definitely having a moment of enormous hype and celebrity investors are interested.
Overall VC funding is down 50% while AI funding is up 100% - We’ve talked about this before. Meanwhile, the only companies pitching VCs seem to be ones who think they have a novel AI technology or novel way to use someone else’s AI technology.
We are now getting some practical numbers behind this. Ashton Kutcher’s Sound Ventures announced they were raising a $243M AI-focused fund, and were able to do so in just 5 weeks. Given everything, this is really fast! There’s certainly an element of “hey, it’s Ashton Kutcher” and who wouldn’t take the call from Ashton Kutcher asking for money? However, we also are in a period of (relatively) high cost of capital, sinking startup valuations, and macroeconomic uncertainty. AI is proving it’s mettle as a buzzy investment that can defy broader trends and getting the term sheets to validate it.
And this money is going out as fast as it is coming in! Not only did Sound Ventures raise a quarter of a billion dollars quickly, they've already spent half of it! OpenAI, Anthropic, and StabilityAI will get ~$120M from the $243M fund. These are big checks and they are going to the name-brand companies in the Generative AI boom. Everyone would be clambering to put money into any one of these companies, and Ashton Kutcher put money into all three.
I make a big deal of the celebrity effect, but it’s really everyone trying to catch the boom. People are raising all kinds of funds for AI investment, ranging from $25M pre-seed funds to $1B venture commitments.
Is the S&P 500 becoming a big bet on AI?
Meanwhile, on the other side of the velvet rope, the public markets are also having an AI moment. 7 of the top 8 components in the S&P 500 have big bets on AI. Nvidia has become an almost pure play on AI; Microsoft and Google are spending billions and their reputations on AI; Amazon (through AWS) is trying to position itself as the ‘bedrock’ of AI platforms; Apple is also developing their own AI tools and the iPhone is the defacto distribution device for many AI tools.
Source: S&P Dow Jones
Altogether, these AI-exposed companies comprise a 23% weighting in the S&P 500.
Figuring out the exact exposure that these companies have to AI is probably impossible. They also are very large companies, so no one business, component, or technology will change them immediately. However:
• Nvidia IS ~66% weighted on AI as that’s about how much of their revenue comes from selling GPUs to data centers
• Google gets ~60% of its revenue comes from Search, but that is a very contested service right now as the gateway for LLMs to the masses is through Search. Google’s ability to sustain this revenue will be heavily dependent on its ability to incorporate AI
• AWS is the major profit driver for Amazon (~75% of operating profit). If their AI strategy pans out, it should enjoy similar margins and be a significant addition to Amazon’s bottom line.
So, if you buy the SPY S&P 500 Index Fund, then you are making a substantial bet (maybe 5%, maybe 10% of your total investment) on the future of AI.
That’s a lot for something that is considered the ultimate diversified investment! AI is a brand new thing and we still largely don’t know where it will go in the next 1, 5, or 10 years. Compare this to other major sectors in the stock market like financial services, manufacturing, real estate, consumer goods, and transportation. Transportation is the new guy here as it’s only been a major industry for about the last 150 years when railroads were big new thing because they allowed your travel from New York to San Francisco in about 7 days. AI is a seedling compared to this, yet it is a comparable driver in the public stock markets.
We are probably in the midst of an AI bubble?
If private markets are really jazzed about AI, and public markets are heavily weighted towards AI, then are we in the midst of an asset bubble around AI investments? Man, I don’t know. But, probably?
VC investment is both ebullient and highly, highly concentrated on AI and specifically, Generative AI. When there is this much investment energy and so few places to channel it, it’s a reasonable conclusion that a bubble-premium is being put on the valuations of companies landing investment. The classic economic theory is:
Too many dollars chasing too few goods → Price Inflation
That’s exactly what is happening in the VC market right now for AI. Funds are encountering almost no obstacles when trying to raise money so long as they invoke their God-given right to spend other people’s money on AI startups. Meanwhile, it may feel like there’s an infinite supply of startups saying “we are the ChatGPT for <insert any industry or job>”, but there is actually an end to that list of companies and and an even quicker end for quality companies. This means there is pressure to inflate valuations for these companies as investors compete to for deals. This is especially true for the brand names like OpenAI, Cohere, Anthropic, Stability AI, Jasper, and Tome.
Source: Yahoo Finance
Meanwhile, public stocks with AI exposure have roared since the current land run began on Nov. 30 2022, when ChatGPT debuted. For example, Nvidia is up ~80% since then while the S&P 500 is up ~4%. This is probably the most extreme example, but not the only one. AI is undoubtedly a great opportunity for these companies, but is it worth an additional $15 billion in net income per year for Alphabet (as implied by a 25% valuation bump since 2022)? Again, I don’t know man. That’s a lot of searches, especially when you consider Microsoft is doing the exact same thing with Bing.
Other Stuff
Google is making deals
Google and Microsoft are in an intense battle to reign supreme on Enterprise AI, but Google is making waves as a deal maker and positioning themselves as an intermediary.
Google and Microsoft are locked in an arms race for AI supremacy, but Google is definitely makings splashes with deal making to build out their "AI Ecosystem". They appear to be positioning themselves as a vital intermediary.
Thus far, they have AI distribution deals with:
• Salesforce
• Box
• UKG
• Dialpad
• Jasper AI
• Canva AI
and AI training/hosting deals with:
• Cohere
• Anthropic
• Midjourney
• AI21 Labs
• Osmo
Maybe they'll make the trains AI passengers ride-on (e.g. the models), maybe not. But they definitely want to be building the train tracks (distribution) and train stations (hosting).
White House Hackathon for the AI Apocalypse
Nobody really knows the potential harms that can be done with-or-to AI today, but the White House wants to find out so they are war-gaming it.
Generative AI is taking more low-wage jobs
Sadly, this is probably just the tip of the iceberg, but AI will take the jobs of a lot of minimum wage workers at Wendy’s. This is Step 6 of creative destruction, so hopefully Step 7 will come soon for these people.
Cyber Security Arms Race
Microsoft’s product teams are probably having a lot of fun imaging the AI vs. AI future of cybersecurity. Like the Pacific Rim movies, except with an IQ above 15.
Lots of AI lawsuits
Having trouble keeping track of all the copyright lawsuits being thrown around in the AI world? Me too. This helps.
Endnote: I don’t have an editor and I do have a dayjob, so please excuse the minor typos and grammatical mistakes.