What Will an #AI Smartphone" Do?

It remains unclear how generative artificial intelligence (large language models) will change our notions of “smartphone” capabilities. One reason is that LLMs might be likened to platforms, operating systems, applications, features or “simply” interfaces between humans and computer resources. 

Though operating systems typically have been considered to manage and control computer resources and interactions, sitting between the hardware and the applications, LLMs might structurally provide similar functions, mediating between the machines and their users. 

That is a role closely related in some ways to the human-to-machine interface, but is much more pervasive in terms of anticipating needs. 

On the other hand, sometimes LLMs might function almost like an app, as when used to support image processing and editing for smartphone cameras or providing on-the-fly translation. In most cases, LLMs might support the features and operation of almost any other software. In such cases the LLM will appear to be a feature or capability. 

Role	Function	Example Use Cases
Operating System	Personalizes and adapts to user behavior and preferences.	- Proactive suggestions for apps, settings, and actions based on context and usage patterns. - Personalized notifications and reminders tailored to individual needs. - Adaptive user interface elements that adjust to user preferences and accessibility needs.
Application	Generates content, automates tasks, and offers creative tools within specific apps.	- Social media: AI-powered content creation (captions, images, even videos) based on user input or prompts. - Productivity: Generating reports, summarizing documents, and drafting emails based on user instructions. - Education: Personalized learning content and practice exercises tailored to individual learning styles and knowledge gaps. - Entertainment: Creating short stories, music, or games based on user preferences and prompts.
Computer-to-Human Interface	Enables natural language interaction and voice control with devices.	- Conversational virtual assistants that understand complex queries and requests, providing relevant information and completing tasks. - Real-time language translation for spoken and written communication, facilitating cross-cultural interactions. - Voice-controlled interfaces that interpret natural language commands and execute actions effortlessly.
Platform	Provides foundational tools and frameworks for developers to integrate generative AI into their apps.	- APIs for tasks like text generation, image creation, and voice synthesis, accessible to developers for app integration. - Pre-trained AI models for specific domains (e.g., healthcare, finance) that developers can leverage in their apps. - Standardized security and ethical frameworks for responsible development and deployment of generative AI on smartphones.
Capability	Enhances existing smartphone features and adds new functionalities using generative AI.	- AI-powered image and video editing tools with filters, effects, and automatic enhancements. - Personalized fitness coaching with workout plans generated based on individual goals and capabilities. - Real-time language translation during phone calls and video conferences, breaking down language barriers. - Smart assistants that manage personal finances, automate repetitive tasks, and optimize daily routines.

All of that will eventually come into play in creating the market for “AI smartphones.” What we do not yet know is how consumers will come to understand what an “AI smartphone” must be capable of doing. 

AI Should Take Advertising to a New Level of Granularity

In principle, artificial intelligence ought to be able to improve advertising precision, taking “personalization” to a new level. 

Consider Disney’s tool called “magic words” that uses AI to match ads with specific scenes in its movies and shows. The tool analyzes and tags scenes to identify content, brands, images, and mood, creating metadata for personalized advertising.

The advantage is better ad targeting, and might be seen as part of a broader move away from “demographics” to “psychographics” in advertising. 

Though audience demographics including attributes such as age, income, and location long have been used to “target” audiences, they paint what would arguably represent only a broad picture of a potential audience. 

Think of all one’s neighbors. Though broadly similar enough to categorize as a single demographic, interests, values and consumer behavior can vary wildly between homes that would appear to be the “same” demographic. Two households might be high consumers of “sports” content, but one household is interested in a variety of sports; the other only in the National Football League, or only one or two teams in the NFL. 

One home has snowboarders or skiers; the other does not. One home spends significant sums on airfares and hotels; the other spends more on recreational vehicles; cars and vehicular vacations. One home owns a boat; the other does not. 

By definition, “mass media” has to target large groups, producing generic ads. The whole point of digital advertising is its ability to target based on interests and relevance “right now.” 

Feature	Mass Media	Digital Advertising
Audience Reach	Broad	Targeted
Targeting	Limited (demographics, media type)	Highly granular (demographics, interests, behavior)
Personalization	Limited (broad messaging)	Highly personalized (individual preferences)
Timing to Potential Buying Behavior	Less immediate (multiple exposures)	More immediate (real-time behavior)
Data and Insights	Limited (individual responses, effectiveness measurement)	Rich (individual responses, campaign optimization)
Cost per Impression	Lower (economies of scale)	Varies (targeting, platform)
Flexibility & Adaptability	Limited (post-production changes)	High (real-time updates)
Creative Potential	Impactful, memorable (storytelling, emotion)	Interactive, engaging (immersive experiences, demos)

Psychographics are about audience values, interests, attitudes, and lifestyles, presumably indicating the motivating values for purchase behavior. In principle, psychographics mean ads can be tailored to resonate with specific needs and desires, making them more engaging and effective.

Psychographics, in principle,  allow for dynamic ad creation, adapting to individual preferences and behaviors in real-time.

Fair Share "Tax" Will be Paid by Somebody: Business Partners and Customers, Most Likely

So-called “fair share” payments by a few hyperscale app providers to internet access providers in Europe means a formal end to network neutrality as well. Recall that the fundamental net neutrality principle is “non-discrimination.” 

At its core, net neutrality means ISPs must treat all internet traffic equally, regardless of the source, destination, content, platform, or application. That has, in practice, meant no blocking, throttling, or prioritizing specific content.

By definition, taxing traffic from a few hyperscale app providers--delivered at the request of the ISP’s own customers--is unequal treatment by source, content, platform and application. 

Granted, the flow of revenue within a value chain can take many forms. But the core principle in the communications business has been that customers pay for their own consumption. The possible new European Union rules on “fair share” shift the revenue flow, allowing ISPs to charge both their own customers as well as a few firms whose products their customers use extensively. 

At a high level, digital infrastructure value flows toward end users and retail customers, while revenue flows from end users back to infra suppliers. 

Content value is more complicated, as are revenue mechanisms. Some professional content creators create value that flows to distributors, and then to consumers. Revenue can flow from end users, advertisers and other business partners towards content creators and distributors. 

E-commerce providers create value that flows up from product creators and suppliers towards consumers. Revenue flows from buyers to sellers and distributors. 

Providers of social media and search functions create value that flows to end users. Revenue flows from advertisers and business partners towards the social media and search providers. 

source: Kearney 

The change in connectivity value might remain relatively unchanged if the EU imposes “fair share” requirements. End users might still be able to access their favored “fair share” apps. But changes always are possible. In the EU, surcharges or fees for some features could develop, as the affected app providers move to maintain their profit margins. 

The affected firms might optimize their platforms to minimize data consumption, which could affect user experience. They could shift resources and investments away from the EU, focusing on markets with less stringent regulations.

The affected companies might raise advertising costs for their partners. 

The affected firms might explore new data monetization methods. Subscription models for specific features or content could develop.

Overall, the affected firms would likely accelerate an exploration of alternative and additional new revenue streams to compensate for the new costs of doing business in the EU. 

Advertising on EU-consumed content might increase. For-fee elements of service could increase. 

The point is that no value chain participant, facing higher costs from one of its suppliers, is going to sit still. At the very least, new efforts will be made to offset the higher costs. 

In principle, the proposed new payments are a tax on doing business in the EU. And, like all taxes, they are simply a cost of business whose costs must be recovered. They will be recovered. And the payment burden will ultimately fall on consumers and business partners.

How Important will Open Source AI Become?

The debate over proprietary versus open source for building artificial intelligence models is a matter of dispute for large language models as it has tended to be in other areas of coding and software development. Meta defended its “open source” approach recently. 

Meta benefits from open source because it “improves our models, and because there's still significant

work to turn our models into products, because there will be other open source models available

anyway, we find that there are mostly advantages to being the open source leader and it doesn't

remove differentiation from our products much anyway,” said Mark Zuckerberg, Meta CEO. 

“First, open source software is typically safer and more secure, as well as more compute efficient to operate due to all the ongoing feedback, scrutiny and development from the community,” he said. “Efficiency improvements and lowering the compute costs also benefit everyone including

us.”

“Second, open source software often becomes an industry standard, and when companies

standardize on building with our stack, that then becomes easier to integrate new innovations into our 

Products,” Zuckerberg added. “Third, open source is hugely popular with developers and researchers.”

That “helps us recruit the best people at Meta, which is a very big deal for leading in any new technology area,” he argued. 

Beyond all that, Meta, as a builder of end user applications, can well take a “layered” approach to business functions. Meta benefits from universal and quality internet access, for example, but it does not have to build all that infrastructure itself. 

And if open source allows Meta to stimulate the creation of digital infrastructure, that helps its own business model, much as universal internet access allows Meta to build its own core businesses. 

Hulu Redux or Something Else for Disney-Warner Brothers Discovery-Fox

If they can work out the details, ESPN (owned by the Walt Disney Company), Fox and Warner Bros. Discovery have reached an understanding on principal terms to form a new joint venture to build a new sports streaming service. 

Those of you who remember what happened with Hulu--originally founded by News Corporation, NBCUniversal and the Walt Disney Company in 2007--might assume the venture ultimately will prove unstable, as the present concept is for each of the firms to own a third of the venture. 

Many of you will predict that management issues ultimately will arise, that interest in participating might wane or that other priorities will convince one or more of the owners to sell their interests. 

Keep in mind that the venture will only license content from each of the owner firms: there will be no transfer of content ownership rights. 

And Disney says it will continue to develop its own branded “ESPN” sports streaming service, while Warner Brothers Discovery will continue to add sports to its “Max” streaming service. On the other hand, some will argue that the new service will represent a sort of “super bundle” of sports programming that could appeal to sports fans who might see the “one service” as preferable to buying three. 

Studies often suggest the “sports enthusiast” portion of the viewing audience ranges from 12 percent to 34 percent of viewers. So it is a significant segment of the audience. 

Study	Methodology	Target Audience	Estimated Percentage	Key Findings
Parks Associates, 2022	Online survey	U.S. broadband households	34%	34% of respondents consider sports "very important" in choosing a streaming service.
Magid, 2022	Online survey	U.S. pay-TV subscribers	18%	18% of respondents said sports are the "primary reason" for keeping their cable subscription.
Ampere Analysis, 2022	Consumer survey	Global markets	20-25%	Estimated 20-25% of global SVOD subscribers watch sports regularly.
Nielsen, 2021	Streaming viewership data	U.S. adults	12%	Sports accounted for 12% of total streaming viewing time in the U.S.
Light Reading, 2021	Survey of industry executives	Global pay-TV market	20-30%	20-30% of pay-TV subscribers are considered "avid sports fans".

Though the new venture might not directly address the cost of sports programming directly, it is among efforts content owners must make to transition to a future where streaming is the primary model, not linear distribution. 

Content Type	Average per-subscriber cost per month (USD)	Range
ESPN	$8.00	$7.46 - $8.54
Other National Sports Networks (e.g., TNT, TBS, NFL Network)	$1.50 - $3.00	Highly variable depending on network and specific rights
Regional Sports Networks (RSNs)	$3.67 - $5.00	Projected to rise. Varies by region, team popularity, and number of teams carried.
Basic Cable Networks (e.g., USA, TNT, TBS)	$0.50 - $2.00	Wide range based on network popularity and content type.
Premium Cable Networks (e.g., HBO, Showtime)	$5.00 - $12.00	Highly variable depending on network and specific content.
Broadcast Networks (e.g., ABC, CBS, NBC, FOX)	$0.00 - $1.00	Primarily advertising-supported model, but retransmission fees charged in some cases.
Streaming Services (e.g., Netflix, Hulu)	$1.00 - $3.00	Varied pricing models based on content library and tiers.

To be sure, this venture could create some economies of scale. If it leads to higher subscriber volume, that could potentially command better deals from rights holders.

It is conceivable the partners might avoid competition for some sports rights and operating savings in marketing might be feasible.

But potential competition from the hyperscalers (Amazon, Alphabet, Apple) would still remain in place, especially as those contestants look to add sports programming.

Microsoft's App Business Aids AI Revenue Model

Compared to Amazon Web Services or Google Cloud, Microsoft benefits from an extensive enterprise software business where AI features can be added rapidly. At least for the moment, that seems to have provided a revenue spark for Microsoft overall, and for Microsoft Cloud operations in particular. 

As Amy Hood, Microsoft CFO said, “strong demand for our Microsoft cloud offerings, including AI services” contributed to the fact that “Azure and other cloud services revenue grew 30 percent and 28 percent in constant currency, including six points of growth from AI services,” Hood said. 

Satya Nadella, Microsoft CEO, pointed out that Microsoft Cloud surpassed $33 billion in revenue and was “up 24 percent,” while Azure AI customers surpassed 53,000. “Over one third are new to Azure over the past 12 months,” perhaps based on the ability to use LLMs from Cohere, Meta and Mistral on Azure without having to manage underlying infrastructure, Nadella suggested. 

Also, Microsoft noted 1.3 million paid GitHub Copilot subscribers, up 30 percent quarter over quarter, and more than 50,000 organizations using GitHub Copilot. 

General-Purpose Vs. Special-Purpose AI

It seems already clear that both general-purpose artificial intelligence models as well as special-purpose AI models will have value. 

Special-purpose models are likely to shine for medical diagnosis, fraud detection, self-driving cars and robotics, for example. 

General-purpose models are likely to continue well enough for personalized assistants, customer service chatbots, content creation, market analysis and scientific discovery.

Special-purpose models for definable use cases will generally have the advantages of high accuracy, efficiency of resource usage and easier documentation of methodology and assumptions.

General-purpose models likely will continue to be more versatile, scalable and arguably more efficient for building some new apps that can use pre-trained models. 

The trade-offs might often include less accuracy for particular vertical (industry specific) or firm applications and might require higher computational resources than special-purpose models already optimized for a single firm, industry or use case. 

But even some who originally saw more promise in special-purpose models seem to be warming to general-purpose models as well. 

One indication of the foundational role of artificial intelligence comes from remarks made by Meta CEO Mark Zuckerberg on that firm’s most-recent earnings call. Zuckerberg says he has changed his mind about how and where to use AI, and the shift is from special purpose and siloed to general purpose. 

“One thing that became clearer to me in the last year is that this next generation of services requires building full general intelligence ,” said Mark Zuckerberg, Meta CEO, on the firm’s most-recent earnings call. “Previously I thought that because many of the tools were social, commerce, or maybe media-oriented that it might be possible to deliver these products by solving only a subset of AI's challenges.”

“But now it's clear that we’re going to need our models to be able to reason, plan, code, remember, and many other cognitive abilities in order to provide the best versions of the services that we envision,” he said.