The "AI apocalypse" panic is overblown
The fear of an AI disaster is the latest technology panic. A popular claim says there is a 10% chance of AI "going off the rails." That number sounds scientific, but there is no real data or calculation behind it. Just because something is a number does not make it a proven fact. There is a lot of room to question that 10%.
The best way for investors to use this panic is through cyber security, not AI directly. As AI gets more advanced, bad actors will use AI to make their hacking stronger.
Palo Alto Networks (PANW)
Palo Alto Networks (PANW) is up more than 100% year to date, almost double from a year ago, and still looks like a buy. Here is the reasoning: even if AI development stopped right now and stayed frozen, hackers would keep getting smarter using the tools already available. They will keep improving how they attack. That is why you need a company like Palo Alto (PANW) guarding the front door.
The nature of hacking has changed. The old model was a burglar smashing a basement window - that is over. Now it is like a burglar ringing your front doorbell in disguise, and you recognize the attacker as a friend and let them in. Firewalls have to be far more complex to stop this. Online scams are more sophisticated: attackers imitate a person's voice so they sound like your family or friend, and fake sites look almost identical to real ones. Telling real from fake is much harder than before.
AI is not slowing down, and regulation would fail
The talk about meta and its AI moves, plus a dinner where heads of major tech companies met President Trump and Chinese President Xi Jinping, shows AI is not slowing down. Sam Altman, Jensen Huang, and Lisa Su of AMD (AMD) were there.
Regulating AI would not work well. Asking the least technically advanced people (regulators) to control the smartest people in the room will fail. It is a very complex industry. New technology historically needs to develop on its own and ends up self-regulated. The 10% threat that scares people is far exaggerated from today's reality.
Regulating AI would be extremely difficult. People also expect too much from how far AI can go. AI does not mimic the human brain at all. On creativity and cleverness, it is fenced in. It works like a "digital mule" that does grunt work we welcome, but it will not gain human qualities, which is the real source of people's fear.
The proof is in the output. AI-made images often get things wrong - the left hand placed where the right hand should be, or a picture that looks nothing like the person even after you supply the exact image and ask it not to change much. This shows AI lacks common sense, which humans have in abundance. As long as AI has no common sense, it will not replace the brain. The underlying technology is completely different. We do not even understand how our own brain works, so we cannot copy it.
Inflection (INFQ) - a quantum computing bet
Inflection (INFQ) is a quantum computing stock that has underperformed the S&P 500 year over year. Quantum computing is hard to understand, and that barrier frightens investors, which is a main reason the stock is down - people are not comfortable with it yet.
Warren Buffett said not to invest in anything you don't understand, but that needs a footnote: we invest in plenty of technology we don't understand, such as meta - nobody but a coder really knows how it works. Quantum computing is the same. It will not replace your laptop, but it can do extremely complex calculations we could never do before, which is what makes it exciting. There are many approaches to quantum computing, and Inflection (INFQ) uses an approach seen as more valuable than the others. Quantum computing could also speed up medical breakthroughs, including for various forms of cancer.
United Rentals (URI)
United Rentals (URI) has had a strong year and still offers opportunity, and it is a relief to think about a non-tech company you don't need deep knowledge to understand. It rents large capital and construction equipment and has been around a long time. It grows by buying and consolidating smaller rental companies - it has done this over 300 times. It now also benefits from data centers: the CEO said on the recent earnings call that data center construction is playing a bigger role. That makes United Rentals (URI) another way to play AI without betting on AI directly. The stock is higher in the pre-market.
Higher bond yields are welcome
This week the 10-year and 30-year Treasury yields hit levels not seen in about 20 years. The 10-year sits at 5.18%, lower by three basis points on the day. Higher yields are a good thing, which is a minority view. For years the 10-year has been unnaturally suppressed. Since COVID we have lived in a near-zero interest rate environment. A 10-year at a 6% or even 7% handle reflects a normal, healthy economy.
If you could see only one statistic in all of capital markets - the 10-year - a near-zero reading over the past 20 years would tell you something is drastically wrong. A very high reading of 18% or 20% would also signal something drastically wrong. A soft trading range of 5% to 8% would not be alarming at all. As long as corporate earnings hold up in that environment, you could even be an optimist and say rates are rising because the country is doing so well.
The counterpoint: on a longer timeline, recency bias shapes how people read today's yields.
