Saturday, October 18, 2025

Wikipedia Pageviews Drop 8%; Generative AI Believed the Reason

The artificial intelligence disruption of existing functions, jobs and businesses continues, with a new bit of evidence coming from Wikipedia


“We are seeing declines in human pageviews on Wikipedia over the past few months, amounting to a decrease of roughly eight percent as compared to the same months in 2024,” Wikipedia says. “We believe that these declines reflect the impact of generative AI and social media on how people seek information, especially with search engines providing answers directly to searchers, often based on Wikipedia content.”


“These declines are not unexpected,” Wikipedia says. “Search engines are increasingly using generative AI to provide answers directly to searchers rather than linking to sites like ours.”


“Many other publishers and content platforms are reporting similar shifts as users spend more time on search engines, AI chatbots, and social media to find information.”


By some estimates, website traffic declines of 15 percent to 70 percent might happen because of AI chatbot substitution. Anecdotally, my own use of Wikipedia has dropped to virtually zero, as I find AI overviews in Google search are an absolutely satisfactory alternative. 


Sector

Possible Change/Disruption

Quantitative Data on Shifts (Where Available)

Notes/Source

Journalism & Media

Automation of article/video generation; reduced human roles; content scraping without payment

Proportion of people believing journalists often use AI up by at least 3 percentage points in 2025; existential threat intensifying financial crisis

https://reutersinstitute.politics.ox.ac.uk/generative-ai-and-news-report-2025-how-people-think-about-ais-role-journalism-and-society 

Creative Industries (e.g., Writing, Design)

Job losses and diminished worker agency; shift to reviewing AI output

300 million global jobs (9.1% of workforce) at risk; 77,999 tech job losses linked to AI in early 2025; 58% decline in tech hiring from 2024 to 2025

https://www.nu.edu/blog/ai-job-statistics/ 

Website/Content Publishers

Traffic diversion via AI search summaries

15-64% decline in organic traffic from AI Overviews; up to 70% drop for some sites; educational platforms saw ~50% decline

https://www.forbes.com/sites/torconstantino/2025/04/14/the-60-problem---how-ai-search-is-draining-your-traffic/ 

Stock Photography

Erosion of licensing revenues; AI generators replacing stock libraries

Stock photo market at $6.1B in 2025 (projected to $11.3B by 2032); AI image gen market from $350M (2023) to $1.1B by 2030

https://colinritman.medium.com/comprehensive-analysis-of-stock-photography-revenue-potential-a-2025-market-assessment-0b881c4d1703 

Music Industry

AI tracks flooding streams; reduced royalties for human artists

AI-generated music could take 20% of streaming revenues by 2028; platforms using AI to lower royalty costs

https://info.xposuremusic.com/article/how-ai-generated-music-could-impact-music-catalog-valuations 

Education (Textbooks, Online Courses)

Shift to AI-personalized materials; reduced demand for static content

No direct loss stats yet; concerns on plagiarism/accuracy prominent; AI reimagining textbooks for adaptive learning

https://www.sciencedirect.com/science/article/pii/S2666920X24000225 

Overall Online Content

AI dominance in creation; commoditization of human work

Estimates suggest 90% of online content AI-generated; 750M apps integrating AI models

https://seedblink.com/blog/2025-04-23-generative-ai-and-the-future-of-content-creation-at-scale 


Friday, October 17, 2025

"My Truth" Versus "The Truth"

The concept of "objectivity" in U.S. journalism education has undergone a significant transformation since the 1970s, and in a more subjective direction, perhaps mirroring the more-subjective intellectual tendencies of post-modern thinking in general. 


Here’s the problem: it has become harder to separate “what a thing is” from “what I say a thing is.” Subjectivism inevitably means there is no external, objective truth independent of individual or cultural perspective. And that is a major problem.


Subjectivism leads to the notion that there is “no such thing as truth.” Instead, there is only a linguistic and cultural creation: “my truth,” as the saying goes. But, by definition, one person’s “truth” (“what is true for me”) cannot be a universal truth shared by all or most other people. 


Such “truth” is essentially “my opinion,” and while that is fine in many situations, it is destructive of broader culture and social cohesiveness. In a sense, it is the logical conclusion when all we focus on is “the individual” and not “society” or “the culture” or the “nation” or the “group.” 


So here’s the problem: When the concept of shared, verifiable facts is undermined, it encourages the belief that all perspectives are equally valid. This makes reasoned compromise and a unified approach to problems virtually impossible, as each side operates from its own constructed, in-group reality or "narrative."


In a nutshell, the problem is that we then cannot commonly agree on “what a thing is,” but only upon what each of us says a thing is. 


This shift toward subjectivism is one of the most defining and consequential trends of post-modern thought, a profound philosophical departure from the Enlightenment-era belief in objectivity, universal truth, and the capacity of reason and science to arrive at a single, verifiable reality. This arguably is most significant in the area of ethics and morality, since there are no “universal truths.”


Field

Modernist Stance (Objective)

Post-Modernist Stance (Subjective/Relative)

Epistemology (Theory of Knowledge)

Knowledge is a process of objective discovery; certain knowledge is possible.

Knowledge is always mediated by individual or cultural context, making it subjective and relative.

Ethics/Morality

Universal moral laws exist (e.g., natural rights, categorical imperatives).

Moral claims are expressions of individual or cultural attitudes (Ethical Subjectivism or Moral Relativism).

History

History is a linear, factual record of events to be objectively uncovered.

History is a narrative or literary form, with no single, objective account; focus is on whose voices (perspectives) have been excluded.

Art/Aesthetics

Art should strive for universal beauty or a coherent, original meaning.

Art is open to infinite, individual interpretation; the meaning resides in the subjective experience of the viewer, not the intention of the artist.


Postmodernism generally views the concept of objective reality as a "social construct" or a product of language and power structures, leading to a focus on the subjective experience.


In journalism, we see a comparable shift from "neutrality" to "truth-seeking," "transparency," and "context," all of which is subjective. 


Since the 1970s, journalism education has increasingly moved away from solely event-centered reporting ("just the facts") to providing analysis, explanation, and context ("what does it mean?"). That might sound like an improvement, but it necessarily is explicitly more subjective. 


And since “absolute impartiality is impossible,” the focus shifts to journalists (in principle) being transparent about their sources, methods, and inherent biases. In principle, journalists are supposed to actively correct for biases. In practice, this often simply means imposing biases. 


Aspect

Prior to 1970s (Traditional "Objectivity")

Post-1970s (Evolving Conceptions)

Guiding Ideal

Detachment and Neutrality (Journalist as a dispassionate recorder)

Truth-Seeking, Context, and Transparency (Journalist as an active interpreter)

Core Practice

"Both-Sides-Ism": Giving equal time and weight to all sides of a controversy.

Verification and Context: Prioritizing factual accuracy, evidence, and scientific consensus over false equivalence.

Journalist's Role

Invisible, impersonal, non-judgmental conduit of "facts."

Self-aware, transparent, providing analysis and interpretation; may involve "moral clarity."

Tone and Style

Simple, straight-laced "inverted pyramid" style; emphasis on the visible facts.

Interpretive, analytical, and narrative-driven; emphasis on the meaning and causes of events.

Critique Addressed

Minimal focus on systemic bias or impact of reporting routines.

Addressing inherent biases, the reinforcement of the status quo, and the silencing effect of traditional neutrality on diverse voices.


The problem is that without a firm commitment to the concept of shared truth, social cohesiveness becomes very hard, and perhaps impossible. Whether in the realm of philosophy, culture, economics or politics, subjectivism is a key problem. 


We cannot fix problems we cannot agree exist. We also can create problems if we cannot agree on some shared understanding of the nature of reality.


In the "post-truth" era, feelings and personal beliefs seemingly matter more than facts.

Wednesday, October 15, 2025

Individuals Working with an AI "Team Mate" Sometimes Outperform Two-Person Teams

A study by a team of researchers finds that AI “significantly enhances performance” perhaps in part by breaking down functional silos. “Without AI, R&D professionals tended to suggest more technical solutions, while commercial professionals leaned towards commercially-oriented proposals,” the authors say. “Professionals using AI produced balanced solutions, regardless of their professional background.”

source: The Cybernetic Teammate: A Field Experiment on Generative AI Reshaping Teamwork and Expertise

Individuals using AI matched the performance of teams without AI, suggesting, at the very least, that AI usage does not harm output. “Our findings show that AI replicates many of the benefits of human collaboration, acting as a “cybernetic teammate,” the team says.

“Individuals with AI produce solutions at a quality level comparable to two-person teams,” they add. “The adoption of AI also broadens the user’s reach in areas outside their core expertise.”

“Overall, our findings indicate that adopting AI in knowledge work involves more than simply adding another tool. By enhancing performance, bridging functional expertise, and reshaping collaboration patterns, GenAI prompts a rethinking of how organizations structure teams and individual roles.”

Tuesday, October 14, 2025

How Big is AI Market? Depends on Your Assumptions

At the moment it is possible to generate an almost-arbitrary “market opportunity” for artificial intelligence, for a number of reasons, including notably the likelihood (virtual certainty) that AI will be embedded into most existing products and services. 


source: Business Insider 


In practice, equity analysts merge hardware, cloud infrastructure, core software, and services into what can be called the “AI technology stack.” That’s the complete infrastructure required to use AI features. Generally, those revenues are about half of total market estimates. 


The other half reflects vertical applications and functional deployments that integrate AI into specific economic sectors, industries and processes. 


Of course, we wind up double counting revenue if we tally “AI” revenues as including all apps, products and services that “use AI” in their value proposition, and then separately enumerate the value of those activities (software of all sorts, automated vehicles, search, social media, e-commerce, for example). 


Revenue Bucket

Representative Products / Activities Included

Key Analysts & Forecast Examples (2025–2035)

Share or Growth Focus by Analysts

Infrastructure Hardware

GPUs, accelerators, AI servers, high-bandwidth memory, edge AI chips, and networking equipment supporting AI compute

Goldman Sachs, Fortune Business Insights, Grand View Research

Base for most AI models; hardware CAGR ~28–30%, driven by hyperscalers (NVIDIA, AMD, Intel)

Cloud and AI-as-a-Service (AIaaS)

AI compute instances, APIs, training & inference services (AWS Bedrock, Azure OpenAI Service, Google Vertex AI)

McKinsey, Fortune Business Insights

70%+ of enterprise AI deployments run on cloud AI environments by 2030

Core Software / Platforms

Machine learning frameworks (TensorFlow, PyTorch), data labeling, MLOps tools, AI model management

IDC, Precedence Research, Grand View Research

Software accounts for ~35–48% of total revenue in 2025

Generative AI Platforms

LLMs, diffusion models, text/image/video generation tools (OpenAI GPT, Anthropic Claude, Midjourney, Runway)

Bloomberg Intelligence, Fortune Business Insights

Estimated $1.3 trillion addressable market by 2033, fastest-growing segment

Professional & Managed Services

Consulting, integration, model fine-tuning, ethical AI audits, and enterprise deployment services

McKinsey, Deloitte, PwC

Among fastest-growing components (CAGR >35%) as firms seek integration expertise

Cybersecurity & Risk AI

AI threat detection, fraud monitoring, anomaly detection, predictive risk analysis

Grand View Research, FinRofca

Included under “function” in future AI workflows; fastest CAGRs in enterprise use cases

Industry-Specific Applications (Vertical AI)

Applied AI in healthcare, BFSI, retail, automotive, manufacturing, law, agriculture

Grand View Research, Precedence Research

Analysts model vertical AI revenues separately (Healthcare, BFSI, Retail largest)

Embedded / On-device AI

Consumer and industrial IoT devices with local inference (smart glasses, phones, vehicles)

Grand View Research (Meta, Apple examples)

Growth tied to edge AI adoption; enables privacy-driven processing

AI Tools for Business Functions

HR tech, marketing, supply chain, operations, finance, customer service (chatbots, copilot-type agents)

Fortune Business Insights, Gartner

Functional AI makes up ~20–25% of total AI revenue by 2030

Quantum AI & Neuromorphic AI (emerging)

Quantum-accelerated AI algorithms, brain-like computing hardware for future models

Fortune Business Insights, McKinsey Technology Outlook

Described as post-2030 “third wave” AI enabler, with gradual commercial integration


A Scenario for Superhuman AI Development over the Next Decade by AI Futures Project

Daniel Kokotajlo, Scott Alexander, Thomas Larsen, Eli Lifland and Romeo Dean, members of The AI Futures Projectproduce a set of scenarios scenarios for "superhuman AI" over the next decade.

 

Daniel Kokotajlo is a former OpenAI researcher. 

Eli Lifland co-founded AI Digest, did AI robustness research, and ranks number one on the RAND Forecasting Initiative all-time leaderboard.

Thomas Larsen founded the Center for AI Policy and did AI safety research at the Machine Intelligence Research Institute.

Romeo Dean is completing a computer science concurrent bachelor’s and master’s degree at Harvard and previously was an AI Policy Fellow at the Institute for AI Policy and Strategy.

Scott Alexander is a blogger and writer. 

They believe the superhuman AI impact "will be enormous, exceeding that of the Industrial Revolution."

Their scenario represents "our best guess about what that might look like."


AI is Solow Paradox at Work

An analysis of 4,500 work-related artificial intelligence use cases suggests we are only in the very-early stages of applying AI at work a...