Motley Fool Money
Episode: "A New Trend in AI is Emerging: Efficiency"
Date: March 26, 2026
Host: Tyler Crowe
Contributors: Matt Frankel, John Quast

Episode Overview

This episode focuses on two core themes:

1. The impact of recent litigation against big tech—specifically Meta and Alphabet—over social media's role in mental health, and how it might (or might not) signal a "tobacco moment" for these companies.
2. A new emerging trend in artificial intelligence: efficiency. The team explores recent developments from ARM and Google that could change the resource and energy demands of AI—and what that means for investors.

The show closes with a mailbag segment tackling the perennial question: Is it better to dollar-cost average with automatic investments, or should you keep cash for "buying the dip"?

Key Discussion Points & Insights

1. Big Tech Litigation & the “Tobacco Moment”

(Starts 00:05)

• Headline News: Recent court verdicts found Meta Platforms (Facebook, Instagram) and Alphabet (YouTube) liable in cases linking social media use to mental health struggles, with punitive damages awarded. Meta lost a separate case in New Mexico involving misleading safety claims.

  • "Meta Platforms and Alphabet were found liable for a woman's mental health struggles related to social media addiction and they had to pay punitive damages." — Tyler Crowe (00:15)

• Personal Impact:

  • Neither Matt Frankel nor John Quast are Meta shareholders; John’s household holds Alphabet shares.
  • Both say these verdicts haven't changed their investment thesis as of yet but acknowledge the landscape is shifting.

• Mental Health Context:

  • Social media's link to mental health is no longer controversial.
    • “Social media apps are engineered to maximize user engagement. I don't think that's a controversial take...it is how they generate revenue.” — Matt Frankel (01:53)
  • Matt recommends The Anxious Generation by Jonathan Haidt for deeper exploration.
  • John, speaking as a parent: "People are becoming a lot more cognizant of the dangers of social media right now than they were a few years ago... My kids, I can tell you, won't be on social media until they're at least in their late teens." (03:27)

• Is This the "Tobacco Moment"?

  • The comparison is discussed extensively—will legislative and legal risk erode the business, as with Big Tobacco?... Or, like tobacco, will tech stocks continue to do well despite total societal awareness of harm?
    • "Even after decades of litigation and declining tobacco use in the U.S., tobacco stocks have still been excellent investments over the years." — Tyler Crowe (04:12)
  • BP (post-Deepwater Horizon) and chemical companies (PFAS/"forever chemicals") cited as examples where legal risk truly did hurt long-term shareholder value.
  • $3M in damages is negligible for these firms, but the precedents could snowball.
    • “Facebook has over 3 billion users worldwide. So would this pave the way for everyone whose life has been harmed by a social media addiction to potentially try to get millions of dollars from these companies?” — John Quast (06:00)
  • The consensus: Too early to tell if this is transformational or not. The cases will be appealed, and the regulatory response is uncertain.

2. AI’s New Inflection: Efficiency Innovations

(Begins 08:48)

• Intro: The conversation shifts from AI's explosive growth and hardware demand to urgent needs for efficiency—energy, memory, and cost.

• Key News Nuggets:

  • ARM’s Move: ARM, known for low-power chips, is pivoting to create its own CPUs under a fabless model. Meta is the first customer (but it isn’t exclusive).
    • “The news that we are getting now is that ARM is actually going to make its own CPUs using a fabless model. So just like NVIDIA or AMD. And Meta is going to be the first customer.” — Matt Frankel (09:24)
  • Google’s Breakthrough: Google announced "TurboQuant," a new way to compress memory for large language models (LLMs), potentially reducing memory requirements by up to 6x.
    • “Google said their recent research showed that a new memory compression method...could potentially reduce the memory requirements of large language models...by a factor of six.” — John Quast (10:13)

• Investor Reaction:

  • Memory chipmakers (Micron, SanDisk) stocks fell; efficiency breakthroughs suggest AI infrastructure may not need as much hardware as previously modeled.
    • "Obviously the knee jerk reaction in the market was for memory companies. Oh, this is bad." — Tyler Crowe (10:46)
  • But efficiency is necessary: physical, power, and supply constraints are making the current trajectory impossible.
    • “These aren't just like good for the AI infrastructure build out. They're absolutely necessary to come anywhere close to meeting the goals and targets that we've been talking about.” — Tyler Crowe (11:04)
  • Power Shortages Impending:
    • "30–50% of data centers in 2026 will be delayed because of power shortfalls...Morgan Stanley projects a 44 GW shortfall through 2028." — Matt Frankel (12:05)
    • Some California centers are built but can’t turn on—grid can't handle it.

• Historical Context & Takeaways:

  • This is the inevitable next step of tech’s evolution—like flat-screen TVs, productivity leaps always follow waves of efficiency improvements.
  • Google’s breakthrough is just for a specific part of memory (“key value cache”), not the entire architecture, so doom for chipmakers is likely exaggerated.
    • "It only optimizes a small part of the memory needs... It's a fraction of a fraction, not a fraction of a whole.” — Matt Frankel (12:50)
  • "Memory stocks like Micron are still up over 300% over the past year, even after the recent pullback." — John Quast (13:10)

• Ultimate Investor Take:

  • Efficiency is an investable theme—look for companies driving energy or compute breakthroughs.
  • "The only way to make all this possible...is we need some breakthroughs in efficiency, in terms of power generation. Something is going to have to happen along the way." — Tyler Crowe (13:47)

3. Mailbag: Automatic Investments vs. Buying the Dip

(Starts 15:21)

• Question:

  • Listener Jay Fung asks whether setting up automated investments is wise, or if "waiting for a dip" makes more sense in the current high market.

• Panelist Approaches:

  • John Quast: Does both: steady auto-investing (dollar-cost averaging, DCA) plus holding some cash to opportunistically buy dips.

    • "It really takes the emotion out of the equation and it mathematically forces you to buy more shares when stocks are cheaper." (16:52)

  • Matt Frankel: Leans towards automatic investment. Cites Fidelity study: investing $5,000 every January 1st from 1980-2023 would generate $5.1M, slightly outperforming monthly DCA ($4.8M). "The earlier you get the money in the market working for you, the better." (18:56)

    • “If you time things perfectly, you did about 10% better than if you had just invested everything on January 1st. And if you time things terribly, you did about 18% worse…Is that 10% upside worth the downside risk of trying to wait for that dip? I don't think so.” — Matt Frankel (20:33)

  • Hybrid Solution Advocated: Put 80-90% on schedule, save 10-20% for major dips so you’re not forced to sell good assets when opportunity knocks.

  • Ultimate Wisdom:

    • "The facts don't change. I'd rather be making my financial decisions based on the facts, because the facts don't change. My feelings aren't reliable." — Matt Frankel (18:24)
    • "Missing out by waiting for the dip is often worse than just being invested—even if the market feels expensive." — Summarized from John Quast (17:30)

Notable Quotes & Memorable Moments

• On Social Media's Addictive Design:

  • "Social media apps are engineered to maximize user engagement. I don't think that's a controversial take...it is how they generate revenue." — Matt Frankel (01:53)

• On the Precedent of Litigation:

  • “Facebook has over 3 billion users worldwide. So would this pave the way for everyone whose life has been harmed by a social media addiction to potentially try to get millions of dollars from these companies?” — John Quast (06:00)

• On AI Power Constraints:

  • "There's massive power requirements. And you've been yelling, 'Great Scott.' And I've been yelling, 'What the heck is a gigawatt?'" — Matt Frankel, joking about the sheer energy needs (11:38)

• On the Need for Efficiency:

  • "These aren't just like good for the AI infrastructure buildout. They're absolutely necessary to come anywhere close to meeting the goals and targets that we've been talking about." — Tyler Crowe (11:06)

• On Market Timing:

  • "If you time things perfectly, you did about 10% better than if you had just invested everything on January 1st. If you time things terribly, you did about 18% worse..." — Matt Frankel (20:33)

• Favorite Ending:

  • "I think the best part of this conversation is I get to use my favorite Charlie Munger quote, which is, 'I have nothing else to add.'" — Tyler Crowe (21:57)

Timestamps for Key Segments

• Big Tech Litigation & "Tobacco Moment": 00:05 – 07:43
• AI Efficiency Innovations: 08:48 – 14:23
• Mailbag: Automatic Investing vs. Buy the Dip: 15:21 – 21:57

The Motley Fool's Takeaway

• Big Tech faces truly novel, potentially precedent-setting legal risks, but it’s too soon to say whether these are existential for their business models or just another cost of doing business, as with tobacco.
• In AI, the “arms race” is now as much about efficiency as scale. Expect pitches around energy, hardware, and algorithm breakthroughs to proliferate—and remember that most big productivity booms come from unseen efficiency gains, not just raw scale.
• For investors, the discipline of regular investing—while opportunistically using dips—has stood the test of time.