Bloomberg Tech — "Tech Left Behind in S&P 500’s Latest Rebound"

Date: December 3, 2025
Hosts: Caroline Hyde (New York), Ed Ludlow (San Francisco)
Episode Summary by Podcast Summarizer

Main Theme

The episode explores why the technology sector continues to lag behind other areas within the S&P 500, despite an overall market rebound. It covers investor anxieties about big tech's future gains and spending, the broadening of market leadership beyond the "Magnificent Seven", surging AI infrastructure demand, custom chip innovation, and the challenges of monetizing generative AI across both software and media industries. Insights from experts, executives, and Bloomberg analysts are woven throughout.

Key Discussion Points

1. Market Overview & Tech’s Lag in the S&P 500

Post-October Diversification (01:21–04:12)
- Investors have been diversifying portfolios after October’s lows; tech underperforms while healthcare, communications, and energy are strong.
- Big names like Nvidia, Apple, and Microsoft are “less magnificent” lately due to volatility and mixed sentiment.
- Ongoing debate on whether tech can continue to deliver outsized gains or if it’s time to seek alpha elsewhere.
- Quote: “As much as we like to think things only go up, the question is can you outperform and is there alpha generation?” — Caroline Hyde (04:04)
Investor Anxiety & Reported Spending Cuts (01:23–02:56)
- Reports (some disputed) that Microsoft is telling salespeople to lower targets for its AI products (CoPilot), fueling concerns about falling productivity gains and tech spending.

2. Breadth Beyond the Magnificent Seven & ETF Flows

Market Broadening & Ex-Mag7 Strategies (05:01–07:40)
- Silvio Jabronski (Defiance ETF CEO/CIO) explains more investors are broadening their exposure beyond the top tech names.
- Flows are rising into ETFs that exclude the Magnificent Seven, recognizing that most investors already have heavy exposure.
- Despite current volatility, there’s optimism technology will remain a market leader long term, especially in AI and infrastructure.

“The average investor has exposure to Mag7 everywhere… so when the market pulls back and you're looking to diversify, we take out those names because we're saying hey, you already have them—hold on to them, but participate in the broadening of the rallies.”
— Silvio Jabronski (06:23)

3. The Evolving AI & Quantum Tech Race

AI Spending & Politics (05:43–07:40)
- AI spending seen as still nascent—impact to widen in coming years.
- Nvidia continues as the “monopoly” in AI chips, but custom semiconductors from competitors are on the rise.
- Political and policy factors in Washington are crucial for market direction, especially for trillion-dollar companies.
China vs. US in Quantum Computing (07:40–09:24)
- Nobel laureates express concern that China is close behind, or even leading, the U.S. in quantum computing.
- American corporations/government tend to ramp up investment whenever they fear being overtaken.

“What happens when we think that, you know, other governments are beating us in AI or technology? The US government tends to invest. And American corporations tend to invest and catch up and try to win the race.”
— Silvio Jabronski (08:42)

4. Volatile Crypto and MicroStrategy’s Strategy

Crypto as a Market Barometer (04:12, 09:24–10:41)
- Bitcoin remains volatile (~92,000), and companies like MicroStrategy face scrutiny over their long-term holding strategies.
- Defiance ETF offers both long and short exposure to such names, mirroring the two-way risk appetite in the current market.

5. Key Company & Executive Interviews

a. AWS & The AI Compute Race (11:41–17:40)

Matt Garman (AWS CEO) Interview
- AWS to double capacity to 8 GW by 2027, expanding both Nvidia GPU and in-house Trainium chip infrastructure.
- Emphasis on customer-driven adoption of custom silicon vs. Nvidia—different chips for different workloads.
- Over half of tokens/inference in Bedrock now run on Trainium.
- Deep partnership with Anthropic (Project Rainier) to help scale next-generation models.
- “[There are] use cases that are best for training and others where Nvidia GPUs are going to be your best option. We want to have all of those available…” — Matt Garman (13:00)

b. Joanne Feeney (Advisors Capital Management) on Custom AI Chips (17:40–22:04)

Amazon’s vertical integration showcases the future: custom-designed chips (like with Marvell) will coexist with Nvidia in a growing “AI compute pie”.
Application Specific Integrated Circuits (ASICs) will win share for tailored AI/ML workloads, but overall demand is so large Nvidia’s dominance will remain for the medium term.
New U.S. government investments (e.g., X Lite’s laser-based lithography, $150mn via Chips and Science Act) aim to bolster domestic semiconductor capability and rival ASML.

“It’s clearly the case that Nvidia is going to lose share over time in the big world of AI compute. But nobody should be surprised by that… the pie is still growing so massively they can’t meet demand.”
— Joanne Feeney (19:18)

c. Dell, Marvell, Okta — Earnings & Infrastructure

Dell: CEO Michael Dell sees continued, insatiable demand for AI infrastructure; says S&P 500 index is a good bet for long-term compounding (25:10).
Marvell: Gains on custom chip contract wins, bullish earnings, and photonics (AI data transmissions with lasers); seen as a sign custom chips are gaining traction (26:46–28:31).
Okta: CEO Todd McKinnon discusses the “massive opportunity” in securing AI via identity management—witnessing unprecedented inbound interest (29:07–33:46).

“All these companies are trying to roll out AI agents... to make these agents useful, they have to be connected to all of a company’s data. If not, the agent can’t work on that data.”
— Todd McKinnon (29:18)

6. AI Content & Social Dilemmas

YouTube Kids and AI ‘Slop’ (36:19–42:45)
- AI-generated content is increasingly being targeted at very young children on YouTube Kids.
- Concerns raised over content quality, the risk to child development and attention span, and the difficulty for parents in managing exposure.
- Parents feel conflicted—many rely on YouTube Kids as a “third parent,” especially during busy times.
- Notable: Australia’s pending social media ban for under-16s exempts YouTube Kids, raising questions about regulatory responses.

“One parent I spoke to described it as a third parent sometimes when she just needs a break.”
— Alex Levine (41:55)

7. TikTok's Massive Brazil Data Center Plans (42:45–44:48)

$37 billion data center, built entirely on clean wind and hydroelectric power, signals TikTok’s major Latin American expansion.
Brazil’s unified national grid and renewable energy leadership are key factors.

8. Salesforce, Microsoft, and Monetizing GenAI Agents (44:48–47:45)

Concerns about monetizing AI copilots and agent products given high inference costs and weak enterprise IT spend.
Even strong agent product launches (Salesforce’s Agent Force, Microsoft Copilot) face skepticism regarding whether they meaningfully move the revenue needle in the near term.

“There is no shortage of revenue on the infrastructure side… but when it comes to copilot products, Office Copilot for example, it’s very expensive to run… Over time that's going to come down and it will lead to more and more people embracing that.”
— Anurag Rana, Bloomberg Intelligence (46:29)

Notable Quotes & Memorable Moments

On Market Anxiety:
“Tech still lagged and we’ve seen it underperform healthcare and other key sectors.” — Host Caroline Hyde (01:23)
On the Breadth of the Recovery:
“400 out of the 500 names actually participated in the bounce… so you see a little bit of breadth expansion there.” — Silvio Jabronski (06:18)
On AI Infrastructure Demand:
“The desire and the hunger for more power and more compute is almost insatiable.” — Matt Garman (12:56)
On Custom Chips Disrupting Nvidia:
“ASICS are very powerful… and are going to be put to use for the type of applications in AI for which they're best designed. So it's clearly the case Nvidia is going to lose share over time in the big world of AI compute.” — Joanne Feeney (19:18)
On AI Content for Kids:
“I was taken aback by the fact that there are so many babies on YouTube and YouTube Kids in the first place… Their brain is still developing, that’s the argument as to why perhaps it should be limited to an extent.” — Alex Levine (36:48)
On the Challenge of Monetizing GenAI:
“All these co pilot products that companies are launching… in the case of Microsoft, when it comes to something like a GitHub copilot, highly productive and massive uptake… Office copilot… is very expensive to run at this point.” — Anurag Rana (46:29)

Timestamps for Key Segments

Market Check & Tech Lag: 01:21–05:01
ETF Flows Ex-Mag7: 05:01–09:24
AWs, AI Chips & Cloud Race (Matt Garman): 11:41–17:40
Custom Chip Development / US Semiconductor Policy: 17:40–22:33
Marvell Photonics & Infrastructure (Dina Bass): 26:40–28:31
Okta, Securing AI Identity (Todd McKinnon): 29:07–33:46
AI & Kids’ Content on YouTube (Alex Levine): 36:19–42:45
TikTok Data Center Brazil: 42:45–44:48
Salesforce & GenAI Monetization (Anurag Rana): 44:48–47:45

Overall Tone & Takeaways

The episode reflects an anxious, data-driven market atmosphere—investors and executives are optimistic about AI’s long-term economic impact, but cautious about near-term earnings, tech sector leadership, and the integrity of both technology and content as innovations accelerate. The tone remains brisk, analytical, and focused on practical implications for markets, business, and society.

Those who missed the episode:

Will grasp why tech is lagging despite big headlines,
Hear from AWS and Okta leaders on the realities and opportunities in AI,
Gain context on custom chips, the quantum race, and ETF strategies,
Understand anxieties about generative AI’s revenue prospects,
Be prompted to think about the social impacts of the AI/media boom.

Bloomberg Tech continues to be a must-listen for anyone tracking the intersection of technology, innovation, and business market dynamics.

Bloomberg Tech — "Tech Left Behind in S&P 500’s Latest Rebound"

Date: December 3, 2025
Hosts: Caroline Hyde (New York), Ed Ludlow (San Francisco)
Episode Summary by Podcast Summarizer

Main Theme

Key Discussion Points

1. Market Overview & Tech’s Lag in the S&P 500

Post-October Diversification (01:21–04:12)
- Investors have been diversifying portfolios after October’s lows; tech underperforms while healthcare, communications, and energy are strong.
- Big names like Nvidia, Apple, and Microsoft are “less magnificent” lately due to volatility and mixed sentiment.
- Ongoing debate on whether tech can continue to deliver outsized gains or if it’s time to seek alpha elsewhere.
- Quote: “As much as we like to think things only go up, the question is can you outperform and is there alpha generation?” — Caroline Hyde (04:04)
Investor Anxiety & Reported Spending Cuts (01:23–02:56)
- Reports (some disputed) that Microsoft is telling salespeople to lower targets for its AI products (CoPilot), fueling concerns about falling productivity gains and tech spending.

2. Breadth Beyond the Magnificent Seven & ETF Flows

Market Broadening & Ex-Mag7 Strategies (05:01–07:40)
- Silvio Jabronski (Defiance ETF CEO/CIO) explains more investors are broadening their exposure beyond the top tech names.
- Flows are rising into ETFs that exclude the Magnificent Seven, recognizing that most investors already have heavy exposure.
- Despite current volatility, there’s optimism technology will remain a market leader long term, especially in AI and infrastructure.

“The average investor has exposure to Mag7 everywhere… so when the market pulls back and you're looking to diversify, we take out those names because we're saying hey, you already have them—hold on to them, but participate in the broadening of the rallies.”
— Silvio Jabronski (06:23)

3. The Evolving AI & Quantum Tech Race

AI Spending & Politics (05:43–07:40)
- AI spending seen as still nascent—impact to widen in coming years.
- Nvidia continues as the “monopoly” in AI chips, but custom semiconductors from competitors are on the rise.
- Political and policy factors in Washington are crucial for market direction, especially for trillion-dollar companies.
China vs. US in Quantum Computing (07:40–09:24)
- Nobel laureates express concern that China is close behind, or even leading, the U.S. in quantum computing.
- American corporations/government tend to ramp up investment whenever they fear being overtaken.

“What happens when we think that, you know, other governments are beating us in AI or technology? The US government tends to invest. And American corporations tend to invest and catch up and try to win the race.”
— Silvio Jabronski (08:42)

4. Volatile Crypto and MicroStrategy’s Strategy

Crypto as a Market Barometer (04:12, 09:24–10:41)
- Bitcoin remains volatile (~92,000), and companies like MicroStrategy face scrutiny over their long-term holding strategies.
- Defiance ETF offers both long and short exposure to such names, mirroring the two-way risk appetite in the current market.

5. Key Company & Executive Interviews

a. AWS & The AI Compute Race (11:41–17:40)

Matt Garman (AWS CEO) Interview
- AWS to double capacity to 8 GW by 2027, expanding both Nvidia GPU and in-house Trainium chip infrastructure.
- Emphasis on customer-driven adoption of custom silicon vs. Nvidia—different chips for different workloads.
- Over half of tokens/inference in Bedrock now run on Trainium.
- Deep partnership with Anthropic (Project Rainier) to help scale next-generation models.
- “[There are] use cases that are best for training and others where Nvidia GPUs are going to be your best option. We want to have all of those available…” — Matt Garman (13:00)

b. Joanne Feeney (Advisors Capital Management) on Custom AI Chips (17:40–22:04)

Amazon’s vertical integration showcases the future: custom-designed chips (like with Marvell) will coexist with Nvidia in a growing “AI compute pie”.
Application Specific Integrated Circuits (ASICs) will win share for tailored AI/ML workloads, but overall demand is so large Nvidia’s dominance will remain for the medium term.
New U.S. government investments (e.g., X Lite’s laser-based lithography, $150mn via Chips and Science Act) aim to bolster domestic semiconductor capability and rival ASML.

“It’s clearly the case that Nvidia is going to lose share over time in the big world of AI compute. But nobody should be surprised by that… the pie is still growing so massively they can’t meet demand.”
— Joanne Feeney (19:18)

c. Dell, Marvell, Okta — Earnings & Infrastructure

Dell: CEO Michael Dell sees continued, insatiable demand for AI infrastructure; says S&P 500 index is a good bet for long-term compounding (25:10).
Marvell: Gains on custom chip contract wins, bullish earnings, and photonics (AI data transmissions with lasers); seen as a sign custom chips are gaining traction (26:46–28:31).
Okta: CEO Todd McKinnon discusses the “massive opportunity” in securing AI via identity management—witnessing unprecedented inbound interest (29:07–33:46).

“All these companies are trying to roll out AI agents... to make these agents useful, they have to be connected to all of a company’s data. If not, the agent can’t work on that data.”
— Todd McKinnon (29:18)

6. AI Content & Social Dilemmas

YouTube Kids and AI ‘Slop’ (36:19–42:45)
- AI-generated content is increasingly being targeted at very young children on YouTube Kids.
- Concerns raised over content quality, the risk to child development and attention span, and the difficulty for parents in managing exposure.
- Parents feel conflicted—many rely on YouTube Kids as a “third parent,” especially during busy times.
- Notable: Australia’s pending social media ban for under-16s exempts YouTube Kids, raising questions about regulatory responses.

“One parent I spoke to described it as a third parent sometimes when she just needs a break.”
— Alex Levine (41:55)

7. TikTok's Massive Brazil Data Center Plans (42:45–44:48)

$37 billion data center, built entirely on clean wind and hydroelectric power, signals TikTok’s major Latin American expansion.
Brazil’s unified national grid and renewable energy leadership are key factors.

8. Salesforce, Microsoft, and Monetizing GenAI Agents (44:48–47:45)

Concerns about monetizing AI copilots and agent products given high inference costs and weak enterprise IT spend.
Even strong agent product launches (Salesforce’s Agent Force, Microsoft Copilot) face skepticism regarding whether they meaningfully move the revenue needle in the near term.

“There is no shortage of revenue on the infrastructure side… but when it comes to copilot products, Office Copilot for example, it’s very expensive to run… Over time that's going to come down and it will lead to more and more people embracing that.”
— Anurag Rana, Bloomberg Intelligence (46:29)

Notable Quotes & Memorable Moments

On Market Anxiety:
“Tech still lagged and we’ve seen it underperform healthcare and other key sectors.” — Host Caroline Hyde (01:23)
On the Breadth of the Recovery:
“400 out of the 500 names actually participated in the bounce… so you see a little bit of breadth expansion there.” — Silvio Jabronski (06:18)
On AI Infrastructure Demand:
“The desire and the hunger for more power and more compute is almost insatiable.” — Matt Garman (12:56)
On Custom Chips Disrupting Nvidia:
“ASICS are very powerful… and are going to be put to use for the type of applications in AI for which they're best designed. So it's clearly the case Nvidia is going to lose share over time in the big world of AI compute.” — Joanne Feeney (19:18)
On AI Content for Kids:
“I was taken aback by the fact that there are so many babies on YouTube and YouTube Kids in the first place… Their brain is still developing, that’s the argument as to why perhaps it should be limited to an extent.” — Alex Levine (36:48)
On the Challenge of Monetizing GenAI:
“All these co pilot products that companies are launching… in the case of Microsoft, when it comes to something like a GitHub copilot, highly productive and massive uptake… Office copilot… is very expensive to run at this point.” — Anurag Rana (46:29)

Timestamps for Key Segments

Market Check & Tech Lag: 01:21–05:01
ETF Flows Ex-Mag7: 05:01–09:24
AWs, AI Chips & Cloud Race (Matt Garman): 11:41–17:40
Custom Chip Development / US Semiconductor Policy: 17:40–22:33
Marvell Photonics & Infrastructure (Dina Bass): 26:40–28:31
Okta, Securing AI Identity (Todd McKinnon): 29:07–33:46
AI & Kids’ Content on YouTube (Alex Levine): 36:19–42:45
TikTok Data Center Brazil: 42:45–44:48
Salesforce & GenAI Monetization (Anurag Rana): 44:48–47:45

Overall Tone & Takeaways

Those who missed the episode:

Will grasp why tech is lagging despite big headlines,
Hear from AWS and Okta leaders on the realities and opportunities in AI,
Gain context on custom chips, the quantum race, and ETF strategies,
Understand anxieties about generative AI’s revenue prospects,
Be prompted to think about the social impacts of the AI/media boom.

Bloomberg Tech continues to be a must-listen for anyone tracking the intersection of technology, innovation, and business market dynamics.

wavePod

Tech Left Behind in S&P 500’s Latest Rebound

Powered by Wave AI

Summary

Bloomberg Tech — "Tech Left Behind in S&P 500’s Latest Rebound"

Main Theme

Key Discussion Points

1. Market Overview & Tech’s Lag in the S&P 500

2. Breadth Beyond the Magnificent Seven & ETF Flows

3. The Evolving AI & Quantum Tech Race

4. Volatile Crypto and MicroStrategy’s Strategy

5. Key Company & Executive Interviews

a. AWS & The AI Compute Race (11:41–17:40)

b. Joanne Feeney (Advisors Capital Management) on Custom AI Chips (17:40–22:04)

c. Dell, Marvell, Okta — Earnings & Infrastructure

6. AI Content & Social Dilemmas

7. TikTok's Massive Brazil Data Center Plans (42:45–44:48)

8. Salesforce, Microsoft, and Monetizing GenAI Agents (44:48–47:45)

Notable Quotes & Memorable Moments

Timestamps for Key Segments

Overall Tone & Takeaways

Summary

Bloomberg Tech — "Tech Left Behind in S&P 500’s Latest Rebound"

Main Theme

Key Discussion Points

1. Market Overview & Tech’s Lag in the S&P 500

2. Breadth Beyond the Magnificent Seven & ETF Flows

3. The Evolving AI & Quantum Tech Race

4. Volatile Crypto and MicroStrategy’s Strategy

5. Key Company & Executive Interviews

a. AWS & The AI Compute Race (11:41–17:40)

b. Joanne Feeney (Advisors Capital Management) on Custom AI Chips (17:40–22:04)

c. Dell, Marvell, Okta — Earnings & Infrastructure

6. AI Content & Social Dilemmas

7. TikTok's Massive Brazil Data Center Plans (42:45–44:48)

8. Salesforce, Microsoft, and Monetizing GenAI Agents (44:48–47:45)

Notable Quotes & Memorable Moments

Timestamps for Key Segments

Overall Tone & Takeaways