All week at the Cut, I’ve been writing to you about the megatrend at the top of everyone’s mind, artificial intelligence – or AI.
We kicked off the week with a look at the blowout earnings report from the world’s No. 1 AI chipmaker, Nvidia (NVDA)… and why it shows the AI opportunity is even bigger than the boldest estimates out right now.
We also heard from colleague Nomi Prins on the three phases of the AI adoption curve.
And our tech investing expert showed why investing in the world’s No. 2 AI chipmaker, Advanced Micro Devices (AMD), has higher upside and lower downside than Nvidia.
All week, you and your fellow readers have been writing in questions. And in today’s mailbag edition of The Daily Cut, Nomi and Colin are standing by with answers.
First up, a question for Nomi…
Reader question: Investing experts keep harping on Nvidia and ChatGPT. Are these the only two options?!
– John F.
Nomi’s response: Thanks for writing in, John. This is a great question.
It’s worth looking at AI-linked stocks in two different groups.
One group is the beneficiaries of AI in the early stages. Their stock prices have quickly shot higher.
Training and deploying AI algorithms requires vast amounts of computational power. That’s why share prices in semiconductor and networking companies such as Nvidia and Supermicro have soared this year.
These companies are providing the computing power, and building networking infrastructure, to get AI off the ground.
Nvidia is up about 240% in 2023. Supermicro, which makes servers optimized for running ChatGPT and other AI systems, is up 235%.
Other early beneficiaries such as Microsoft, Google, and other software companies are adding AI into their apps and search engines.
This is where you need to be careful with bubble-like moves in stock prices and valuations.
But while there may be a bubble in some AI stocks, we’re not in a bubble in AI overall. The full impact of AI has yet to be felt by the industries it will change the most.
Tech companies are getting all the attention right now from investors who want to profit from AI. And they’re the most obvious play. But some of the most disruptive potential of AI will be felt in banking, biotechnology, and energy.
For example, banks will use AIs to streamline their lending process. Over time, these systems will become better loan officers than their human counterparts. They can examine vast datasets of what makes make a creditworthy loan and apply those insights to new loan creation.
In biotech, AIs are speeding up the drug discovery process. In fact, one of the greatest achievements of AI so far is folding more than 200 million proteins.
Predicting the shapes proteins will fold into in our bodies is key to the drug creation process.
It typically takes a PhD candidate the full length of a four- to five-year PhD program to figure out the shape of a single protein.
But AlphaFold, a protein-folding AI, can fold an average-length protein in seconds. This allowed it, in 2021, to fold 200 million proteins in a single year.
It then put that out a free-access database that researches from all over the world can use.
And they’re using it to create vaccines against malaria, combat antibiotic resistance, develop new medicines, and shed light on ways to slow the progression of Parkinson’s, Alzheimer’s, and other neurological diseases.
Researchers are also using AIs to create more stable plasma fields to contain nuclear fusion reactions. This is one of the biggest roadblocks to successful fusion and a future of limitless renewable energy.
And there’s a lot of time left for AI’s full potential to play out. You just have to be selective about which stocks you buy, in which sectors, and not overpay for the stocks that have already moved.
That’s why I just shared details on my favorite AI stock right now in an urgent briefing called The AI Ultimatum.
It’s trading for $0.26. And a major announcement could send it soaring.
I won’t get into the details here. You can watch my briefing in full here.
Nomi isn’t the only member of the team who’s excited about the potential of AI. So is our tech investing expert, Colin Tedards.
Colin is a coder turned investor. He heads up our tech investing e-letter The Bleeding Edge where he’s made it his mission to help as many readers as possible profit from the rise of AI.
He says we’ll see an AI bubble inflate, and inflate, and inflate as AI changes just about every aspect of our lives. That means potentially life-changing gains for folks who invest accordingly.
Reader question: I’ve been to seances where the table answered questions while communicating with the dead. Was that okay? Is it okay to contact the spirits? Just because we can? No. I put AI in the same category. It’s not necessary for a fulfilling life, and it’s dangerous. I’ll drive my own car, thank you.
– Sid C.
Colin’s response: Sid, thanks for the message. There’s going to be a lot of discussion around AI and its impacts. As there should be.
As with all new technology, there will be some uses that aren’t for the better. And it will be easy to focus on these and lose sight of the benefits AI will bring.
But all technologies come with trade-offs. They make our lives easier and more efficient. But they mean change. And that can make us uncomfortable.
Let me offer one more potentially spooky implication of AI… Soon, it will be possible to raise the dead.
An AI can be trained on all written and spoken material from a historical figure… or from a relative.
Combine this with realistic generative images (deep fakes), and we’ll be able to speak with them.
You’re right, the power of this technology is going to feel uncomfortable at times. But AI will democratize access to knowledge in ways we’ve never seen before. AIs will also be huge timesavers.
Time is the most valuable resource we have. AI will give us more of that.
And lastly, a question for Colin about all extra energy we’ll need to power AIs…
Reader comment: With the hype and projections surrounding AI, I don’t hear much about where all the electricity will come from to power them.
EVs, robots, blockchains, and supercomputers require a constant supply of juice. Factories with robots can run 24hrs a day all year long. This means the machines they operate will run all day as well. That’s a 4x drain on our current power supply right off the bat.
– Roger T.
Colin’s response: Roger, thanks for writing in. This is a forward-thinking observation.
In many ways, the country that solves the energy challenge wins the AI race. I’m not confident Washington will move fast enough. So innovation will likely fall to the private sector.
Data centers are energy hogs. In 2020, they were estimated to consume about 200 terawatt-hours of energy per year globally. That’s roughly equivalent to the annual energy consumption of Thailand, a country of 71 million people.
Data centers use huge amounts of water for cooling. Some large data centers can use as much water in a year as 100,000 average U.S. households.
And data centers running AI computations are the most resource-intensive tasks. That means AIs consume even more energy and water.
At some point, we’ll have an event or a series of news articles that force the data centers to be more proactive about power and water management.
Nike and Apple have experienced this as it relates to poor working conditions in manufacturing facilities. We’ll have this moment as it relates to data centers soon.
Luckily, these large companies are flush with cash. So, with enough public pressure, the innovation will occur.
I’m looking at several companies to recommend in this area. There are several new technologies that cut energy and water use in data centers. I look forward to bringing you those recommendations very soon.
That’s all we have time for today. Remember, you can write in your comments, questions, and concerns for any of the Legacy Research team at feedback@legacyresearch.com.
Have a great weekend.
Chris Lowe
Editor, The Daily Cut