Technology Does Not Eliminate Human Choice

Technological determinism is a recurrent idea, suggesting that technology is a primary force shaping society. The notion is that once a technology appears, its capabilities and economic logic tend to push society toward particular outcomes. But a survey of Goldman Sachs interns suggests the limits of such thinking. 

About 60 percent of respondents prefer no AI involvement in creative work such as art, music, and books. They are comfortable with some AI assistance for content curation, healthcare, shopping, customer service or personal finance, so long as there is human validation. 

source: Goldman Sachs 

In other words, the survey suggests a human response to AI that is not a simple or passive role, despite much “doomerism” about it. 

To be sure, technology creates possibilities. But that does not necessarily mean it is imperative. The telephone could have eliminated face-to-face meetings, but it didn't. Television could have eliminated newspapers, but it didn't. ATMs could have eliminated bank tellers. The automobile made commuting possible but did not make commuting inevitable.

Calculators changed which mathematical skills were economically valuable, but not lower-order operations. When computation became cheap, computation itself became less economically scarce. The value shifted toward applications, data, organizational knowledge and decisions.

The point is that predictions about deterministic technological outcomes have often been wrong. 

Technology	Deterministic prediction	What actually happened	What it illustrates
Telegraph	Near-instant communication would effectively eliminate the importance of distance and perhaps produce a unified world	Distance became less important for information, but geography, politics, culture and institutions remained extremely important	Technology can reduce one constraint without eliminating others
Telephone	Telephones would largely replace face-to-face interaction	Telephone communication became complementary to physical interaction	People value how interaction occurs, not merely information transfer
Radio	Mass broadcasting would create a common national culture and potentially make older media obsolete	Radio became one medium among many; audiences fragmented and new media emerged	New media usually complement as well as substitute
Television	Television would replace newspapers, radio, books and perhaps traditional education	All survived, although their roles changed substantially	Technologies rarely eliminate an entire human activity simply because they can perform part of it
Automobile	Cheap personal transportation would make cities and mass transit obsolete	Cars produced suburbs and highways, but dense cities, rail and walking remained valuable	Technology interacts with urban design, policy and preferences
Air conditioning	Climate would cease to constrain where people live and work	It transformed settlement patterns, but water, energy, geography, culture and economics continued to matter	Removing one constraint doesn't remove the system of constraints
Nuclear power	Extremely cheap electricity would make energy essentially unlimited and transform society	Nuclear power became important but expensive, politically contested and institutionally constrained	Technical potential ≠ economically or politically realized potential
Computers	Automation would eliminate large portions of clerical and administrative employment	Computers eliminated many tasks but also created enormous new categories of work and increased demand for information-intensive occupations	Productivity and employment aren't mechanically inversely related
ATMs	Banks would need dramatically fewer employees because machines could replace tellers	ATM deployment coincided with changes in branch economics, but banks retained many tellers and expanded some branch networks	Lower cost per transaction can increase total demand
The Internet	Physical retail, newspapers, television and other intermediaries would disappear	They were disrupted, but many adapted; new intermediaries emerged	Disintermediation often produces reintermediation
E-books	Paper books would rapidly disappear	E-books became substantial but physical books remained remarkably persistent	Consumers don't optimize solely for technical efficiency
MP3/digital music	Recorded music would become essentially free and traditional music companies would disappear	Streaming created new subscription and platform models; recorded music remained a major business	New technology changes the business model rather than necessarily destroying the activity
Social media	Universal connectivity would produce greater understanding and democratic consensus	It also produced polarization, fragmentation, communities and entirely new forms of political communication	Technology amplifies social forces rather than determining their direction
Online education	Internet video would make physical universities largely unnecessary	Online education grew enormously, but residential universities remained important for social, credentialing and institutional reasons	Education is an institution, not merely information delivery
Remote work	Once high-quality videoconferencing existed, offices would largely become unnecessary	Remote/hybrid work expanded substantially, but many organizations retained physical workplaces	Work has coordination, social and cultural dimensions
3D printing	Manufacturing would move from centralized factories to individuals producing objects at home	3D printing found important niches, but mass manufacturing remained highly centralized	Unit economics and supply chains matter alongside technical capability

Technology does constrain the choice set, though. Once something becomes dramatically cheaper, faster or more capable, people change their behavior. 

Social media reshapes content production, consumption and value. Automobiles changed the geography of cities.

Artificial intelligence can automate many routine operations. But efficiency is not the only key value for humans. People don't optimize every aspect of life for minimum cost, greatest efficiency or least time. 

Often, value is generated by activity itself. Live music, handmade objects, sports, religious services, face-to-face education, crafts, gardening, cooking and all manner of “live” experiences have intrinsic value, so we keep doing those things. 

So the issue is more “what do you do with more free time?” 

AI might allow us to focus more on purpose, meaning, responsibility, relationships and judgment as the need to spend so much time on instrumental activities decreases. As unfashionable as “religion” might be in some quarters, it is possible its value could become more important. 

Humans make choices, even when technology creates possibilities or conditions. We don't always choose efficiency over other values.

If Internet Was About "Disintermediation," AI "Remediates"

If disintermediation removing distributors in the value chain) was the hallmark impact of the internet on media and content, artificial intelligence arguably reinserts a new mediation layer in the value chain. 

If the internet tended to remove gatekeepers, AI is going to bring that function back.

The internet let creators, publishers, and businesses reach users directly.

• Authors could self-publish.

• Musicians could bypass record labels.

• Retailers could sell directly online.

• Newspapers could distribute without print trucks.

The dominant idea was that gatekeepers lost power and relevance.

AI seems to be inserting a new form of mediation between users and content, especially as agents proliferate:

• Users ask an AI instead of visiting websites

• Buyers rely on AI agents instead of browsing stores

• Students consult AI instead of textbooks or tutors

• Professionals ask AI instead of reading reports.

If the internet reduced friction between producer and consumer, allowing creators to reach consumers directly, AI inserts a new layer the AI agent). 

The practical consequences:

• Traffic shifts from websites to AI interfaces.

• Brands lose direct contact with customers.

• Original creators may receive less attribution.

• AI becomes the gatekeeper.

Still, both the internet and AI function in some similar ways:

• Reduce transaction costs

• Increase access

• Disrupt existing business models

• Favor large platforms. 

What is different is the way AI affects the monetization of content. Instead of search engine optimization, AI abstracts the sources of content.

Where SEO was intended to deliver relevant links, AI “gives you the answer.”

During the internet era, media companies lost control over distribution.

During the AI era, they may lose control over:

• Discovery

• Attribution

• Monetization

• Customer relationships

The risk is greater because AI can both create and summarize content, reducing the need to visit the original source.

But something greater than “distribution” or “discovery” is at work. Where the internet largely focused on “discovery” of content, AI agents will essentially “create” content.

As the internet forced content providers to change, so AI will force changes as well. Organizations must optimize not only for human customers, but also for AI systems.

Era	Core Trend	Central Dynamic
Internet	Disintermediation	Remove middlemen
Search	Aggregation	Organize information
Social Media	Platformization	Concentrate attention
AI	New intermediation	Machine agents become the middlemen

The shifts are from discovery to meaning; traffic volume to citation; site visits to inclusion in summary; active search to passive consumption. 

Feature	Internet Era Shift	AI Era Shift
Primary Value	Providing access to information.	Providing synthesis/meaning from information.
Core Metric	Clicks and Traffic.	Conversions and Authority/Citations.
Publisher Role	Destination (user visits your site).	Source (AI summarizes your work).
Advertising	Targeted banner/link ads.	Integrated conversational/agentic ads.
User Experience	Active searching and navigation.	Passive, conversational consumption.

The internet was about removing traditional gatekeepers. AI seems destined to insert a new layer into the value chain.

Participants will have to adapt. 

The primary challenge for publishers and content providers is “zero click.” When a user gets their answer directly on the search results page, they have no reason to visit the publisher’s site, leading to significant declines in organic traffic and ad impressions.

Advertisers now must design for agentic discovery as they once designed for search engines.

Platforms are performing a delicate balancing act: they need to keep users satisfied with quick, AI-driven answers while ensuring that content creators (the "web") don't disappear, as the AI needs that content to function.

Metric	Old Search Model	AI Search Era
Primary Goal	Traffic volume (clicks)	Brand authority & Citations
User Journey	Linear/Multi-click	Compressed/Conversational
Measurement	SEO rankings/CTR	Conversion-based/Brand impact
Publisher Value	Being a destination	Being a trusted source

One often hears lamentations about the unfairness of the new system, typically in the form of who gets the value from creating and distributing content. But nothing in the media and content businesses is guaranteed. Nobody has a “right” to audiences. 

New gatekeepers are coming, whether we like it or not.

Now Claude Managed Agents Threaten Edge Networks

Add edge networks such as Akamai, Fastly and Cloudflare to the list of enterprise software of service providers that might be disrupted or helped by growing use of language models and especially agents. 

And blame Anthropic for the recent concern. Claude Managed Agents, launched April 8, 2026 in public beta, handles sandboxed code execution, checkpointing, credential management, scoped permissions, and end-to-end tracing.

Investors seem to be pricing in only danger, but upside arguably also exists.

The core tailwind is simply volume. Cloudflare's CEO noted that if a human were shopping for a digital camera and might visit five websites, an AI agent doing the same task might visit 5,000. 

In fact, AI bot traffic across the Akamai network surged more than 300 percent since Akamai began tracking it, with AI training crawlers accounting for the largest share. 

One might point to earlier concerns about Akamai capital spending that has rattled equities of many types, but that does not seem to be the impetus for the declines of April 10. 

Fastly has noted that inference operations likewise are providing a big boost to its revenue. 

So why did Akamai stock sell off sharply, about 17 percent, on the news? Cloudflare sold off 13.5 percent on April 10. And Fastly sold off nearly 22 percent on the same day. 

Simply, Claude Managed Agents threatens to move much of that former “outsourced” or external traffic “in house.” Claude is now arguably getting into the infrastructure layer for agents, as a potential competitor. 

Impact area	Mechanism	Direction	Effect on CDNs	Who benefits	Source
Explosive growth in bot & agent traffic	Claude Managed Agents spawn long-running sessions that browse the web at machine speed — the Cloudflare CEO noted agents may visit 1,000× more sites than a human doing the same task. Akamai tracked a 300% rise in AI bot traffic across its network in 2025.	tailwind	More requests, bandwidth, and cache-hit pressure — CDNs must handle vastly more non-human HTTP volume.	All three; Cloudflare leading	TechCrunch
Edge inference & AI Gateway demand	Anthropic routes MCP tool calls through third-party services; CDNs like Cloudflare offer AI Gateway products that proxy and cache LLM API calls at the edge. Workers AI inference requests grew 4,000% YoY in Q1 2025.	tailwind	New high-margin inference and gateway revenue streams open up for CDNs that offer edge compute.	Cloudflare (Workers AI, AI Gateway); Fastly (semantic caching)	Klover.ai
MCP server hosting & orchestration	Claude Managed Agents connects to external services through MCP servers. Cloudflare extended its Workers platform with MCP Server Portals; Fastly offers edge orchestration for MCP-style workloads.	tailwind	CDNs become the natural 'control plane' for routing agent tool calls with low latency and built-in auth.	Cloudflare (MCP Server Portals); Fastly	Cloudflare blog
Bot management & agent identity	Agents generate traffic that is hard to distinguish from malicious bots. Akamai and Cloudflare must evolve bot management to allow legitimate agent traffic while blocking bad actors. Cloudflare & GoDaddy partnered specifically to separate trusted agents from scrapers.	tailwind	Increased demand for bot management, agent identity verification, and WAF rule updates.	Akamai (Bot Manager); Cloudflare (Bot Fight Mode)	The Register
DDoS & security complexity	AI agents can be weaponised or exploited. Cloudflare mitigated a 31.4 Tbps DDoS attack in Q4 2025; hyper-volumetric attacks grew 700% that year. Managed agent traffic is a growing attack surface.	tailwind	Higher security service revenue; CDNs are critical infrastructure for absorbing AI-amplified attack volumes.	All three; Cloudflare most exposed	Cloudflare 2026 report
Semantic caching of agent queries	AI agents issue repetitive or semantically-similar requests. Fastly's 'AI Accelerator' uses semantic caching to avoid redundant inference passes. Cloudflare AI Gateway adds response caching across providers.	tailwind	New caching product category; CDNs monetise the gap between raw inference cost and cached delivery.	Fastly (AI Accelerator); Cloudflare (AI Gateway caching)	Futuriom
Anthropic runs its own agent infrastructure	Claude Managed Agents runs on Anthropic's own infrastructure — sandboxed containers, state management, and orchestration are all handled in-house. This keeps a large slice of compute spend inside Anthropic's cloud, not CDN PoPs.	headwind	Enterprises that previously self-hosted agents (and used CDN edge compute for inference) may shift workloads onto Anthropic's platform instead.	Cloudflare Workers AI; edge compute providers	The New Stack
Disintermediation of traditional CDN caching	AI agents retrieve answers from Claude rather than browsing cached web pages, reducing human-driven page requests. If AI mediates content discovery, fewer end-user requests reach CDN-cached origin servers.	headwind	Long-term erosion of traditional web caching revenue as human page views decline relative to bot/agent queries.	Akamai (most exposed, hardware-intensive legacy CDN model)	FinancialContent
AI-driven content disengagement	Akamai observed agentic traffic declining post-Cyber Week 2025 because many sites lack the structured data agents need. Agents disengage from unoptimised sites, shifting traffic away from poorly-structured CDN customers.	headwind	CDN customers in retail/media face traffic shifts if they don't optimise for agent consumption, reducing effective CDN utilisation.	Akamai customers in retail & media	Akamai AI Pulse
Capital expenditure burden (Akamai-specific)	Akamai is investing heavily in physical GPU inference infrastructure (Akamai Inference Cloud). CapEx is projected at 23–26% of revenue in 2026, weighing on margins even as AI-adjacent revenue grows 45%.	mixed	AI agents drive new revenue but at the cost of high capital intensity for hardware-focused CDNs — software-defined competitors benefit more.	Akamai (headwind); Cloudflare (tailwind — software model)	FinancialContent
Agentic content monetisation standards	Akamai predicts that 'agentic commerce will scale only where trust and permission are clearly established.' Agent identity frameworks (Visa Trusted Agent Protocol, Skyfire KYA) will require CDN-level enforcement of who agents are and what they can access.	mixed	CDNs that build agent identity and permission infrastructure early win new platform revenue; laggards become 'dumb pipes'.	All three — early movers win higher-margin platform roles	Akamai AI Pulse

So even if some tailwinds exist, Claude Managed Agents threaten to transform some of those tailwinds into headwinds. 

The key headwind is that Anthropic runs Managed Agents on its own cloud infrastructure. Users simply describe the agent, and Anthropic handles sandboxing, authentication, and tool execution. 

This keeps substantial compute spend inside Anthropic's own stack rather than at CDN edge nodes.