AI is Going to be Used, Despite Energy or Water Impact

Despite some protestations, it does not appear that artificial intelligence capabilities--no matter what the impact on water or energy consumption; threats to privacy; intellectual property issues; teaching methods or other social and economic effects--are going to remain unused, anymore than the internet was destined to remain unused, or electricity, or the internal combustion engine or any other general-purpose technology. 

GPTs simply are too transformative, affecting most to all economic or social activities. And though we could well be wrong, artificial intelligence seems poised to provide more leverage than prior GPTs. 

source: Ark Investment, CMC Markets 

source: Ark Investment, CMC Markets 

So energy and water impact, for example, have to be kept in context. Yes, AI is likely to increase consumption of both of those resources. But gains might vastly outstrip those inputs, with some measure of lessened resource impact for activities that can incorporate AI. To the extent AI automates and proactively reduces inputs, that should result in footprint reductions elsewhere in the economy. 

The point is that if an observer does not particularly care about economic impact; use of domestic sources; national interest considerations; the cost of energy or impact on lower-income and middle-income consumers, an argument for limiting AI uses might make sense. 

If AI leads to higher water and energy inputs, and if all one cares about is reducing water and carbon footprint, the actual choice of energy sources might not matter much. What matters is simply reducing consumption. 

As a practical matter, all those concerns do matter, even if some energy sources perform better than others in terms of carbon or water impact, and even if some sources have greater or lesser externalities of other sorts (impact on wildlife, land use, conservation of other natural resources). 

Few policymakers would accept a lower water and energy footprint--no matter what--if the tradeoff is lower living standards for working and lower-income citizens; lower economic growth and higher overall costs of living. 

Energy Source	Energy Efficiency	Carbon Footprint	Water Footprint
Nuclear Energy	High	Low (no direct greenhouse gas emissions)	High (for cooling)
Hydropower	High	Low (no direct greenhouse gas emissions)	Moderate (can impact ecosystems)
Solar Power	Moderate-High	Low (no direct greenhouse gas emissions)	Low
Wind Power	High	Low (no direct greenhouse gas emissions)	Low
Biofuel	Moderate	Can vary (depends on type and production method)	Moderate
Coal	Low	High	High (for mining and cooling)
Natural Gas	Moderate-High	Moderate	Moderate

Water and energy footprint is a problem we need to work on, with a view to total benefits and costs, even if some inputs grow because we use AI.

"Platforms" Change Power Relationships in Value Chains

In the linear video business, there long has been a running debate over where “value and power” lie in the value chain, even if both content and distribution matter. 

At times, it has been argued that “content is king,” (while at other times it seems “distribution is king” (Netflix now; Comcast at the height of linear distribution).

But these days, across many industries, it might be the case that “platforms” have blurred the older distinctions between content and distribution as value drivers, across media, software and internet value chains. 

Era	Value	Key Characteristics
1950s-1970s	Distribution	• Limited TV channels controlled distribution • Networks dominated viewership • Content creators reliant on networks for distribution
1980s-1990s	Content	• Rise of cable TV increased channel options • Premium content providers like HBO gained power • Hit shows became more valuable
2000s	Distribution	• Cable/satellite providers consolidated • Bundling gave distributors leverage • Limited streaming options
2010-2020	Content	• Streaming services proliferated • Netflix, Amazon, etc. invested heavily in original content • Bidding wars for popular shows and creators
2021-Present	Hybrid	• Content remains crucial but very expensive1 • Distribution platforms also important as market saturates2 • Streaming services focus on profitability and subscriber retention

Those strategic differences arguably also have applied during the internet era, when at times internet access (distribution) has appeared to drive more value, and other times when apps or services have arguably driven more value. 

The twist for the internet value chain is the role of “platforms,” which do not neatly fit the “content versus distribution” dichotomy. 

Platforms have blurred the lines between content creation and distribution, making the distinction less clear-cut than in previous eras. Many platforms now act as both content creators and distributors. For example, Netflix and Amazon Prime Video produce original content while also distributing third-party content. 

Social media platforms such as YouTube, TikTok, and Instagram also raise different questions about the value of content versus distribution. Facebook’s platform both enables user-generated content, but also provides the value of distribution, simultaneously. 

In fact, the whole disruption of most value chains by the internet was precisely “disintermediation,” which removed some distributors from value chains (e-commerce disintermediated retail stores). 

And platforms (marketplaces) now essentially create new forms of distribution that are advantageous for many asset owners (owners of short-term lodging assets), compared to legacy providers that are displaced (hotel chains). 

Era	Value	Key Characteristics
1990s	Distribution	• Limited internet access • Dial-up modems and ISPs controlled access • Content creation tools were limited
Early 2000s	Content	• Broadband expansion • Rise of blogging and user-generated content • Google's focus on quality content for search rankings
Mid-2000s	Distribution	• Consolidation of ISPs • Net neutrality debates • Rise of social media platforms as content gatekeepers
2010-2015	Content	• Smartphone proliferation • App stores democratized content distribution • Netflix and streaming services invested in original content
2015-2020	Hybrid	• Content remained crucial but very expensive • Platform algorithms gained importance • Rise of influencer marketing
2020-Present	Content	• Increased demand for digital content during pandemic • Growth of creator economy • AI-powered content creation tools

The sorts of dynamics might be inferred for the software business as well, where at different times it has seemed that “apps” drive value, where at other times “distribution” arguably drives more value. 

Era	Value	Key Characteristics
1980s-Early 1990s	Distribution	• Limited distribution channels (retail stores) • Software companies reliant on publishers • High barriers to entry for independent developers
Mid 1990s-Early 2000s	Content	• Rise of the internet as a distribution channel • Emergence of shareware and freeware • Increased accessibility for independent developers
Mid 2000s-2008	Distribution	• Consolidation of major software companies • Dominance of Microsoft in operating systems and office software • Rise of enterprise software suites
2008-2015	Content	• Launch of app stores (Apple App Store, Google Play) • Explosion of mobile apps and indie developers • Democratization of software distribution
2015-2020	Hybrid	• Content remained crucial but discoverability became challenging • App store algorithms gained importance • Rise of subscription-based software models
2020-Present	Content	• Increased demand for specialized software solutions • Growth of no-code/low-code platforms • AI-powered development tools empowering creators

The key observation is that older dynamics (content versus distribution) still exist, but within a context where platforms now are increasingly important gatekeepers and value generators. At the physical layer, data centers and transmission networks and internet service providers remain “distribution.”

At the business and revenue layer, platforms have emerged as the key drivers of value (Facebook social media, Amazon e-commerce, Google search, Uber ridesharing, and all those embody value creation roles similar in some ways to “content” and also similar in fundamental ways to “distribution.” 

That explains why it is easier to say that “hybrid” models now predominate. For “hybrid,” substitute the word “platform.” 

Platforms such as YouTube, Facebook, TikTok and Spotify might be thought of as intermediaries--neither traditional content creators nor distributors--that aggregate content from various creators and curate it. But in some ways the platforms also act as distributors of that user-generated content. 

So “platformization” more accurately modifies former content and distribution functions in some ways, while displacing each of those functions in some ways, as exemplified by “direct to consumer” sales models. 

Perhaps the best example is Amazon, which, in its function as an e-tailer, connects buyers and sellers. In that role it essentially replaces the traditional distributor (retail stores). But Amazon also curates and funds the creation of actual content as the provider of the Amazon Prime streaming video service. 

In its role as the provider of Amazon Web Services “computing as a service,” Amazon acts as a software supplier in its own right as well as a distributor of those products. 

But if we really have to choose, we’d likely argue that platforms have usurped distributor roles more than app and content provider roles, even in instances where the platforms do a bit of both. 

Where AI Will Drive the Greatest Value Chain Impact

Virtually every observer might agree that artificial intelligence will automate laborious tasks and therefore increase process efficiency. AI should also accelerate decision making, as it enables rapid information processing. 

AI should enable more personalization than already is possible for user interactions and experiences and as a byproduct could change the nature of work, entertainment and learning. 

Generative AI, though, might bring cost impact in different ways than did other computing innovations. Virtually all computing eras since the advent of the personal computer have led to lower marginal costs of doing things. 

PCs meant computing power itself was widely available to people. The internet attacked the cost of sharing information and communicating while cloud computing arguably reduced software distribution costs while boosting the ability to apply accumulated data and insights more widely in real time. 

The mobile era extended computing capabilities “everywhere” and untethered from desks, tables or laps. 

Era	Computing Paradigm	Marginal Cost Implications
PC	Personal Computing	- High upfront costs for hardware and software - Relatively high marginal costs for upgrades and maintenance - Limited scalability
Internet	Networked Computing	- Reduced costs for information sharing and communication - Increased accessibility, but still significant infrastructure costs - Marginal costs tied to bandwidth and server capacity
Cloud Computing	On-Demand Computing	- Significantly lower upfront costs - Pay-as-you-go model reduces marginal costs - Improved scalability and flexibility - Potential for cost optimization through resource management1
Mobile	Ubiquitous Computing	- Lower device costs compared to PCs - App-based ecosystem with low distribution costs - Increased connectivity, but data costs can be significant
Future AI	Intelligent Computing	- Potential for near-zero marginal costs in some applications - High initial investment in AI development and infrastructure - Continuous learning and improvement may reduce long-term costs2

So it is reasonable to ask what the AI impact will be, especially generative AI, which seems to be driving mass market and most business AI use cases. 

Angela Strange, Andreessen Horowitz general partner and James da Costa Andreessen Horowitz partner, specialized in enterprise and business-to-business software, including financial technology. 

They believe the AI era leads to lower marginal cost of client and customer interactions, using AI agents to reduce the cost of labor involved in many customer support operations, including those involving information retrieval (files, ledger entries, past transitions, billing and account status). 

source: Andreessen Horowitz 

As applied in many areas outside of financial technology, the value of generative AI is squarely on its impact on content creation. 

Whether we look at text, image, video or audio, GenAI seems destined to have the highest impact on any process or industry built on content creation and its distribution or consumption. GenAI will be useful in any number of customer support contexts, but might be impactful in financial terms for the production of software and code; entertainment content; education and training; business communications; many types of research; marketing and sales.