Maybe AI Doesn't Kill "Search"

One reason Alphabet’s equity valuation has been muted recently, compared to some other “Magnificent 7” firms, is the overhang from potential antitrust action, which has been partially addressed in the United States v. Google decision by Judge Amit Mehta, where the ruling does not require structural separation of the Chrome browser from the rest of Alphabet. 

The other problem is the concern that the search business model could be disrupted by language model chatbots. 

We might not know much about some other antitrust cases involving Google for some time. 

But we might already be seeing signs that Alphabet’s innovations around integrating chatbot functionality with search are paying off. 

“We know how popular AI Overviews are because they are now driving over 10 percent more queries globally for the types of queries that show them, and this growth continues to increase over time,” said Sundar Pichai, Alphabet CEO. 

The point is that we still do not know the longer-term changes in either search or chatbot business models, especially since agent capabilities are coming. That will blur the functions of “learning, research or finding” with “taking action.

Feature	Classic Google Search	Chatbots	Google with Agentic AI
Research (Info Retrieval)	Yes	Yes	Yes
Personalized Recommendations	Yes	Yes	Yes
Multi-step Task Execution	Limited	Limited	Yes
Real-world Action Capability	No	Rarely	Yes (book, buy, schedule, etc.)
Trust & Security	High	Varies	High (Google brand)
Ecosystem Integration	Extensive	Siloed/fragmented	Extensive

Both chatbots and search engines will be able to retrieve information and act on it. The perhaps classic example is trip planning, which blends task-oriented activities and research.

 

But add agent capabilities and the “search” platform becomes the “take action” platform. For a firm such as Alphabet, that means the value of Google search as an advertising revenue generator is augmented by e-commerce revenue, something Google already does to an extent with product searches.

 

The point is that Google search has a plausible path to surviving and perhaps even outshining chabots also outfitted with agent capabilities. 

Somes "Fiber" is Not the Answer

If you have spent any time at all thinking about the cost of building next-generation access networks, you know that customer density really matters. Where rural fiber deployment costs range from $3,000 to $6,000 per household passed, for the most part, urban costs are $700 to $1,500 per household passed, primarily because of customer density. 

Very-low population density (less than 10-20 households per square mile) is one challenge. But challenging terrain also matters, which is why fixed networks are so scarce in mountainous regions, for example. 

To be sure, government subsidy programs can help, but there are always going to be some locations in the United States, for example, that simply won’t be logical candidates, even with subsidies. One reality is that about six percent of the U.S. land mass is “developed” and relatively highly populated. 

Conversely, about 94 percent is unsettled or lightly populated, including mountains, rangeland, cropland and forests.

source: USDA

So it is possible that AT&T might find that up to 15 percent of its copper access lines cannot be upgraded for fiber, and will have to use some form of wireless access.

Verizon, with a denser customer footprint, might ultimately find that up to 10 percent of its access lines, likewise will have to use a wireless platform. 

Lumen, with the least-dense territory of all the former Regional Bell Operating Companies, could easily find that 20 percent to 30 percent of remaining copper access lines cannot be upgraded for fiber. 

The point is that there are physical reasons, beyond customer preferences, for using fixed wireless, mobile wireless or satellite for access to customers in remote and thinly-populated areas.

How Big a Market for "Direct to Phone" Services Such as SpaceX Wants to Provide?

The purchase by SpaceX of EchoStar  AWS-4 and H-block spectrum licenses marks a beginning and an end. The end is EchoStar’s hopes of becoming a fourth facilities-based mobile service provider in the U.S. market.

The beginning is SpaceX’s ambition to become a facilities-based provider of satellite “direct to mobile phone” (D2M) service in the U.S. market. It remains unclear how big that market could become, and some will doubt it is a full-on substitute for conventional mobile service. 

Low estimates by Juniper Research might suggest global revenues of  $30 million in 2025 scaling to $1.7 billion  by 2029, assuming the U.S. market represents 60 percent of the global market. 

Grand View Research estimates are higher, forecasting global revenues of as much as $435 million in 2025, with U.S. share at about 31 percent.

MarketsandMarkets suggests a 2025 market of $560 million in 2025, with U.S. at 42 percent share). 

Those estimates include direct-to-phone services that are either exclusive (as SpaceX will provide) or as supplementary service to a terrestrial mobile service for which an extra charge is required. 

Year	Low Estimate (Juniper-based, $M)	Mid Estimate (Grand View, $M)	High Estimate (MarketsandMarkets, $M)	Average Estimate ($M)
2025	18	135	235	129
2026	45	181	324	183
2027	113	242	447	267
2028	283	324	617	408
2029	708	434	851	664
2030	N/A	581	1,174	918

Compare that to estimated total U.S. mobile operator revenue in 2025 ranging from $340 billion to $344 billion. Satellite direct-to-phone is going to be a niche, though it will be an important niche for users who cannot get cell service otherwise, in hard-to-reach areas (mountains, very-rural areas not alongside major highways, on the oceans, in disaster areas where the mobile network is inoperable). 

The transaction provides SpaceX with valuable mid-band spectrum assets—AWS-4 covering 2000-2020 MHz downlink and 2180 MHz uplink, and H-block spanning 1915-1920 MHz uplink and 1995-2000 MHz downlink. The key is that those frequencies allow satellite signals to be used by standard smartphones. 

Such D2M services could enable seamless texting, calling, and low-bandwidth browsing during natural disasters or for off-grid users, without requiring specialized hardware like external antennas. 

Speeds might start modest (2-4 Mbps), but the value lies in ubiquitous global coverage.

D2M could serve as a complementary "always-on" layer, filling coverage gaps that affect about 20 percent to 30 percent of the U.S. land area where people normally are found.

A Stunning "Controlled Experiment" Shows How Hard "Good Public Policy" Can Be

Students of the effectiveness of public policy might agree, if they thought about it, that we rarely have any way of assessing the effectiveness of policies we devise to solve stated problems, such as protecting students and teachers from Covid infections, for example. 

But sometimes we accidentally do find the closest proxy to a controlled experiment that we are ever likely to find in real life. 

During the COVID-19 pandemic, particularly in the 2020-2021 school year when many public schools in the United States were closed or operating remotely due to lockdowns and health concerns, the majority of Catholic schools remained open for in-person or hybrid instruction. 

According to data from the National Catholic Educational Association, 92 percent of Catholic schools were offering full-time in-person or hybrid learning by early 2021, with most reopening in the fall of 2020 where local regulations permitted. 

This contrasted sharply with large public school districts in cities like New York, Los Angeles, and Chicago, which often stayed closed for the entire year. 

Regarding COVID-19 transmission, there were no reports of major widespread outbreaks or significant spikes in cases attributable to Catholic schools during this period. 

Multiple sources, including diocesan reports and analyses, indicate that infection rates in these open Catholic schools were generally low and comparable to or below community averages, thanks to rigorous health protocols. 

For instance, in the Archdiocese of Washington, officials reported in late 2020 that safety measures were effective even as regional cases rose, with no evidence of school-driven surges. 

Similarly, a 2023 analysis noted that Catholic schools maintained in-person learning without significant increases in infection rates, despite 17 percent of teachers being high-risk. 

Broader studies on school reopenings, such as those tracking U.S. districts in 2020-2021, found that in-person instruction in elementary and secondary settings (including private ones) did not lead to substantial community transmission when mitigation strategies were in place. While isolated cases occurred. Catholic schools were frequently highlighted as models for safe reopening, with no documented "major problems" like mass closures due to outbreaks. 

Title/Publication	URL	Publication Date
Why Catholic schools didn't fail while public ones did during COVID (New York Post)	New York Post	November 15, 2022
While public schools closed on fear and politics, Catholic schools opened on love and science (Washington Examiner)	Washington Examiner	June 28, 2023
Catholic Schools Have Stayed (Mostly) Open During The Pandemic (WGBH)	WGBH	February 25, 2021
COVID heroes: Catholic schools safely reopened (USA Today)	USA Today	February 12, 2021
COVID-19 cluster size and transmission rates in schools from ... (PMC)	NIH	N/A (academic study, circa 2022)
NCEA Releases 2020 - 2021 Data on State of Catholic Schools ... (NCEA)	NCEA	February 2021
ED624233 - Catholic School Enrollment Boomed during COVID ... (ERIC)	ERIC	2022
Analysis of Catholic School Reopening Plans in Fall 2020 (Digital Commons @ LMU)	Digital Commons	2021

It seems fairly obvious now, with our better understanding of Covid transmission, Covid vaccine effectiveness and lethality for people of different ages and health states, that the policies of closing public schools was a mistake. A rather big mistake.

New AI Browsers Can Embrace Different Content Partner Revenue Models

New or different revenue models for content partners are possible when apps such as Perplexity’s 

Comet Plus adds new revenue models for use of the platform. Now priced at $5 per month, Comet Plus has established a revenue pool to shareComet Plus subscription free with content providers. 

Perplexity will keep 20 percent of revenue, while the remaining 80 percent will be distributed to participating publishers.

Partner revenue is earned when:

• a user clicks on a link within the Perplexity app to visit a publishing partner's website.

• when Perplexity's AI-generated summary cites a piece of content from a publishing partner.

• when Perplexity's AI assistant, or "agent," visits a publisher's website to perform a task on behalf of the user.

This is a departure from past search engine policies, and is enabled by the existence of a subscription revenue source. 

OpenAI and Google content licensing deals, for example, often use a one-time or multi-year lump sum payment for the right to use the publisher's content.

The new Perplexity model is directly tied to user engagement. 

Also, the model for ad-supported search engines has been sharing advertising revenue. Publishers are compensated through traffic and subsequent ad revenue generated on their own sites from that traffic. That becomes an issue when AI engines simply provide direct answers and reduce the need for users to click through to content provider sites. 

Perplexity's model explicitly addresses the "zero-click" problem of AI summaries and compensates publishers even when a user doesn't click through, as long as the content is cited. That raises a new issue for publishers, namely how to increase chances their content is cited. 

Infra Suppliers and Mobile Operators Might Need 6G More than Users

We can expect to continue hearing quite a lot from suppliers and service providers about the speed, capacity advantages and compelling new use cases for 6G mobile networks. There are obvious reasons for such claims. 

Simply put, every internet access network has required more capacity over time, and mobile services are not exempt from this need. But there are other reasons for the chatter. 

Suppliers have to convince their customers to replace the existing platform (5G) with the next-generation platform. 

Service providers, on the other hand, have to convince regulators to release new spectrum to support the "revolutionary" new networks. 

And claims about important new use cases are typically part of the argument, aside from the virtually-certain improvements in speed (bandwidth or capacity boosts of 10 times have been normal for each mobile platform since the time of 2G) and lower latency. 

Still, at some point, quantifying the value of a new high-speed network for a single user is a challenge. Beyond a certain point, and that point is nearly always far lower than many expect, faster speeds do not provide any measurable improvement in application performance. 

That noted, It's also impossible to calculate the return on investment of an application that has not yet been invented or is not yet possible because the networks will not support it.

Before 4G, no one could have predicted the rise of on-demand ride-sharing. And while mobile video streaming had been noted as a 3G network innovation, those networks did not generally have the capacity to support such apps at scale. So capacity sometimes does matter.

Still, ride-sharing might have blossomed more because of the use of smart phones, plus GPS, rather than "more bandwidth" as such. 

And while 4G bandwidth improvements did make video streaming usable in nearly all cases, one also has to point to the value of social networks and entertainment as the drivers of value for consumers. And it has generally proven difficult to predict which innovations will emerge at scale for any given next-generation mobile network.

Text messaging emerged for 2G networks almost by accident, for example, as the text capability was developed as a user service after the new Signaling System 7 was adopted for network operations.