Skip to main content

Our approach to advertising and expanding access to ChatGPT

Title: Our approach to advertising and expanding access to ChatGPT Author: Unknown Type: press_release Published: 2026-01-16

Task 1: Agency and Accountability Audit​

About this task

This task examines how agency—the power to act, decide, and be held accountable—is linguistically engineered within the text. It identifies mechanisms (passive constructions, nominalizations, personified abstractions) that manage perception of who acts and who is acted upon, classifying strategies as erasure, delegation, diffusion, inversion, collectivization, or personification.

Naturalizing Technological Evolution Through Personification​

Quote: "AI is reaching a point where everyone can have a personal super-assistant"

  • Participant Analysis: The subject 'AI' is the actor. Human developers, engineers, and corporate strategists who determine the trajectory of the software are absent.
  • Agency Assignment: Agency is delegated to the technology itself. The grammatical structure suggests AI is an autonomous entity evolving of its own volition.
  • Linguistic Mechanism: Personification and abstract actor.
  • Agency Strategy: Personification
  • Power Analysis: By portraying AI as 'reaching a point' naturally, OpenAI evades responsibility for the specific design choices—like adding ads—that they are making.
  • Counter-Voice: We have designed our AI to include a personal super-assistant and an advertising delivery system.

Erasing Human Intervention in Algorithmic Sorting​

Quote: "Answers are optimized based on what's most helpful to you."

  • Participant Analysis: 'Answers' is the goal/patient; the optimizer is deleted. The 'you' is the beneficiary but lacks control.
  • Agency Assignment: Agency is obscured through the passive voice. The human/corporate criteria for what constitutes 'optimal' are hidden.
  • Linguistic Mechanism: Passive voice with agent deletion.
  • Agency Strategy: Erasure
  • Power Analysis: This prevents users from questioning who defines 'helpful'—whether it serves the user's needs or OpenAI’s engagement and revenue targets.
  • Counter-Voice: Our engineers program the algorithm to prioritize specific answers that we define as helpful.

Obscuring Corporate Growth as Democratic Giving​

Quote: "Today we’re bringing Go to the U.S. and everywhere ChatGPT is available, giving people expanded access"

  • Participant Analysis: 'We' (OpenAI) are the agents; 'people' are the recipients. The action is 'giving,' which implies a gift rather than a transaction.
  • Agency Assignment: Agency is explicit but framed as altruistic. The commercial nature of 'expanding' into a new market is reframed as an act of generosity.
  • Linguistic Mechanism: Transitive verb with positive semantic prosody ('giving').
  • Agency Strategy: Inversion
  • Power Analysis: It positions the company as a benefactor rather than a market actor, making the introduction of monetization (ads/subscriptions) seem like a secondary, supportive act.
  • Counter-Voice: We are expanding our market reach to the U.S. to monetize a larger segment of users.

Diffusing Responsibility for Economic Disparity​

Quote: "Who gets access to that level of intelligence will shape whether AI expands opportunity or reinforces the same divides."

  • Participant Analysis: 'Access' (a nominalization) is the subject. The socio-political actors who control that access are absent.
  • Agency Assignment: Agency is diffused into an abstract concept of 'access.' It makes inequality seem like a byproduct of technological flow rather than policy and pricing.
  • Linguistic Mechanism: Nominalization and abstract actor.
  • Agency Strategy: Diffusion
  • Power Analysis: This framing allows the company to present its 'diverse revenue model' (ads) as a solution to inequality while ignoring how their own pricing structures created the divide.
  • Counter-Voice: Our decisions regarding pricing and ad-placement will determine who can afford high-quality information.

Task 2: Ideology and Common Sense Audit​

About this task

This task audits lexical choices, identifying where seemingly neutral words smuggle in contested values, assumptions, or hierarchies. It examines what worldview a phrase wants the reader to accept as "common sense" and explores alternative framings.

Framing Surveillance as 'Personalization'​

Quote: "—and can turn off personalization if you want."

  • Lexical Feature Type: Euphemism / Metaphorical framing

Ideological Work: This choice naturalizes the idea that tracking is a 'personal' service for the user's benefit rather than a mechanism for data extraction and sale.

Inclusion/Exclusion: Positioning the user as having 'choice' makes the extractive system seem voluntary and benign.

Alternative Framings​

PhrasingWorldview CenteredMakes Visible
"Ad-tracking and data profiling"Privacy/CriticalThe extractive nature of the data collection process.
"Targeted marketing filters"CommercialThe intent to manipulate consumer behavior.

The Meritocratic Myth of the 'Level Playing Field'​

Quote: "AI tools level the playing field even further, allowing anyone to create high-quality experiences"

  • Lexical Feature Type: Metaphor / Cultural model

Ideological Work: It invokes the American myth of meritocracy, suggesting technology alone can erase structural economic disadvantages.

Inclusion/Exclusion: It masks the fact that those with the most capital can still leverage these tools more effectively than 'anyone.'

Alternative Framings​

PhrasingWorldview CenteredMakes Visible
"Standardizing content production through automation"Labor/EconomicThe devaluation of human craft and the centralization of tools.
"Lowering the barrier for corporate imitation"SkepticThat 'high-quality' here refers to professional aesthetic, not intrinsic value.

Reifying Intelligence as a Commodity​

Quote: "Who gets access to that level of intelligence..."

  • Lexical Feature Type: Reification

Ideological Work: By calling it 'intelligence' rather than 'software' or 'data processing,' the text elevates the product to an essential human quality that must be managed.

Inclusion/Exclusion: Normalizes the commodification of cognitive processes.

Alternative Framings​

PhrasingWorldview CenteredMakes Visible
"Who gets access to these computational predictions"Scientific/MaterialistThat AI is a processed output, not a sentient quality.
"Who can afford the premium search results"Class-basedThe pay-to-play nature of the information environment.

Task 3: Positioning and Solidarity Audit​

About this task

This task analyzes how texts construct social positions and relationships between speaker and audience, power-holders and the powerless. It examines the implicit "we" and "they"—who is positioned as authority, who as complicit, who is erased.

Quote: "We’ve been working to make powerful AI accessible to everyone"

  • Positioning Mechanism: Inclusive pronoun 'We' (Corporate-as-Collective)
  • Relationship Constructed: A sense of partnership between the corporation and the global public. It positions OpenAI as a steward of a common resource.
  • Whose Reality Wins: The corporate reality—where profit-seeking is synonymous with public service—is naturalized.
  • Power Consequences: It makes resistance to OpenAI's business model seem like resistance to 'progress' for everyone.

Positioning the User as an Autonomous 'Decider'​

Quote: "You control how your data is used. You can turn off personalization..."

  • Positioning Mechanism: Direct address / Imperative mood
  • Relationship Constructed: A false sense of symmetry. The user is told they are in control, while the platform sets all the underlying parameters.
  • Whose Reality Wins: The neoliberal reality where individuals are responsible for managing their own privacy within predatory systems.
  • Power Consequences: It shifts the burden of privacy and ethical consumption onto the user, absolving the company of structural accountability.

Establishing Authority through Register of Transparency​

Quote: "Given that, we want to be clear about the principles that guide our approach..."

  • Positioning Mechanism: Formality and 'Principle-based' register
  • Relationship Constructed: Teacher-student or Ethical leader-follower. OpenAI defines the moral ground before the user can react.
  • Whose Reality Wins: The corporation's internal 'principles' are presented as the objective framework for the entire industry.
  • Power Consequences: It pre-emptively frames ads not as an intrusion, but as a principled implementation of a mission.

Task 4: Discourse Strategies​

About this task

This task identifies overarching strategic patterns—the key moves that the text makes to accomplish its ideological work. Each strategy must cite instances from Tasks 1-3 and articulate material consequences.

The Democratization of Commercialization​

Cited Instances: Naturalizing Technological Evolution Through Personification, Obscuring Corporate Growth as Democratic Giving

Linguistic Patterns: The combination of altruistic verbs ('giving,' 'accessible') with personified technology ('AI is reaching a point') and the erasure of financial agents.

Ideological Function: It frames the expansion of an advertising-based business model as an ethical necessity for 'closing the divide.'

Material Consequences: It justifies the extraction of user data and attention from lower-income tiers ('free and Go') as a philanthropic 'opportunity.'

Counter-Discourse: A discourse that identifies ads as a 'tax' on those who cannot afford privacy.

Ethical Shielding​

Cited Instances: Manufacturing Consent Through the Inclusive 'We', Establishing Authority through Register of Transparency

Linguistic Patterns: Usage of high-level abstractions like 'Mission,' 'Principles,' and 'Humanity' to justify specific technical and financial shifts.

Ideological Function: It prevents the questioning of the profit motive by wrapping it in the language of existential mission (AGI).

Material Consequences: Ensures that the platform's transition into an ad-network is seen as consistent with its 'Charter,' protecting its brand value during a pivot.

Counter-Discourse: A discourse that separates corporate 'charters' from legally binding consumer protections and profit requirements.

Task 5: Structural Mystification Audit​

About this task

This task applies three Critical Theory concepts:

  • Reification (Lukács): Social relations appear as natural objects
  • Social Amnesia (Jacoby): Historical struggles are systematically forgotten
  • False Separation (Adorno): Structural issues framed as individual problems

Part A: Reification Analysis​

Intelligence as a Natural Force​

Quote: "AI expands opportunity or reinforces the same divides."

  • Reification Mechanism: Social processes (hiring, education, capital flow) are attributed to the abstract noun 'AI' as if it were a weather pattern.
  • What's Obscured: The specific investment decisions, labor conditions, and legal frameworks that actually determine wealth distribution.
  • De-Reification: Recognizing that 'divides' are reinforced by how humans price and restrict the tools they build, not by the tools themselves.

Part B: Social Amnesia Analysis​

The Erasure of the Non-Profit Origin​

Quote: "our pursuit of advertising is always in support of that mission"

  • What's Forgotten: The original founding promise of OpenAI as a non-profit specifically intended to be free from financial pressures like advertising.
  • Function of Amnesia: By 'forgetting' that the mission was once incompatible with ads, the text makes the current ad-pivot look like a natural evolution rather than a reversal.

Part C: False Separation Analysis​

Individualizing Systematic Privacy​

Quote: "You control how your data is used... you can clear the data used for ads at any time."

  • False Separation: Structural data harvesting is framed as a series of individual 'choices' (opt-outs).
  • What's Actually Structural: The underlying architecture of the LLM and the ad-network requires mass data collection to function; individual deletions do not change the model's training or the company's data-driven power.
  • Ideological Function: It makes the user feel like an agent while they remain a data-point in a system they did not design.

Synthesis: How These Mechanisms Work Together​

Reification, amnesia, and false separation converge to present OpenAI's commercial evolution as an inevitable, natural, and ethical event. By reifying 'Intelligence' as a force, the text masks the human labor and capital behind it. By inducing social amnesia regarding its original non-profit ethos, it prevents the public from seeing the current move as a breach of trust. Finally, by separating the 'user's choice' from the 'platform's structure,' it privatizes the social problem of surveillance. Together, these strategies conceal a totality where cognitive tools—once framed as a public good—are being enclosed within a private, ad-supported infrastructure. This prevents a collective consciousness from demanding non-commercial, public-interest AI models.

Conclusion: Stakes and Counter-Discourse​

About this section

This section synthesizes the analysis: naming the ideology, tracing material stakes, and providing counter-discourse examples.

Ideology and Material Stakes​

The core worldview constructed here is 'Technological Solutionism'—the belief that social inequality and human progress are best managed through private corporate tools rather than public policy or democratic oversight. The political project is the normalization of 'Surveillance Cognition,' where every interaction with one's own thoughts (mediated through AI) becomes a moment for data extraction and commercial interruption. Materially, this serves the interests of OpenAI's investors by creating a sustainable, scalable revenue stream from the attention of billions. The harm lies in the enclosure of the digital commons; when 'intelligence' is tiered and ad-supported, the quality of information available to the poor becomes fundamentally different (and more manipulated) than that available to the rich. This makes collective action against tech monopolies difficult because the monopoly successfully positions itself as the only provider of 'broad access' to the future.

Counter-Discourse Principles​

A resistant discourse must first practice 'De-reification' by replacing 'AI' with 'OpenAI's proprietary software,' making the actors visible. It must employ 'Historical Memory' by reminding readers that this platform was funded and built on the promise of being a public benefit, not a commercial ad-engine. 'Structural Thinking' would involve connecting the need for ads to the specific demands of venture capital rather than 'humanity's' needs. Finally, it must 'Center the Marginalized' by highlighting how the 'Go' tier targets those who cannot afford to pay for privacy, effectively creating a 'privacy divide' under the guise of 'closing the intelligence divide.'

Counter-Discourse Examples​

OriginalReframedStakes Shift
"AI is reaching a point where everyone can have a personal super-assistant."OpenAI has developed software that can automate many tasks, which we are now deploying as a commercial product.Shifts the focus from a natural technological 'arrival' to a deliberate corporate 'deployment,' making the designers accountable.
"You control how your data is used."OpenAI grants users limited settings to hide certain data points while maintaining the platform's overall extractive architecture.Exposes the power imbalance; 'control' is revealed as a limited set of permissions granted by the more powerful party.
"Ads can be transformative for small businesses."Our ad-platform requires small businesses to pay us for visibility in an environment we have consolidated.Reveals the rent-seeking nature of the platform rather than the 'transformative' benevolence.

Extended Processing Summary​

Gemini's Processing Steps
What you are about to read…

Gemini refers to the text below as "thought summaries." This is an overt consciousness projection because "intentions" are hallmarks of a conscious mind that "knows" what it is doing and why. The concealed mechanistic process is probabilistic text generation. Treat this as just another rhetorical artifact—a way of making the model's processing legible. When the model writes "I should double check that...", it is mathematically altering its own probability distribution for the next token to be more accurate. The "I" is a functional gear in the machine, not a self-reference. This language is a deliberate interface design choice, not a biological reality. The developers could have programmed the model to output passive logs. Instead, they chose to simulate a narrator.

Framing the Announcement

I'm currently engaged in analyzing the recent OpenAI announcement. My focus is on the agency and accountability aspects, initially. I am also working on identifying the ideologies presented, and assumptions that are being framed as "common sense". This process is structured using a Critical Discourse Analysis framework with specific tasks to guide the investigation.

Unpacking Discourse Strategies

I'm now diving into the discourse strategies employed in the announcement, analyzing their impact. I've pinpointed the core theme: introducing ads as "expanding access." Key actors are OpenAI, the user, AI, small businesses, and humanity. I'm focusing on linguistic techniques like passive voice, metaphorical framing of ads, inclusive "we", and reification. The transition to a commercial entity is obscured.

Identifying Persuasion Techniques

I am now closely examining the specific rhetorical strategies at play. "Answers are optimized..." avoids agency, obscuring the "who" and "why." The "mission" statement diffuses accountability. Framing a negative as a positive ethical stance with "we do not optimize..." is also central. I am now deep in reviewing the euphemisms, semantic prosody, and metaphor, too.

Defining Key Ideologies

I'm now identifying core ideologies at play. "Access" is being used to justify ads, creating a false equivalence with democratization. Moralizing the market is also evident in mission alignment. I'm also examining the reification of "AI" and its implication. I'm taking notes on social amnesia related to the company pivot, false separation of user data/privacy, and the stakes of commodified cognition. Explicitly naming surveillance, extraction and profit is now included.

Finalizing Analysis Framework

I am now structuring my findings using a JSON format, organizing by agency, ideology, and structural mystification. Specifically, I am drafting instances, refining them for clarity. I've pinpointed AI personification, erasure through passive voice, and ideological framing. I'm also addressing the market's moralization, and the implications of treating "Intelligence" as a commodity for synthesis. The final polish focuses on the transition.

Refining Ad Impact Analysis

I'm now zeroing in on the implications of an "ad-supported medium." I'm examining how this framework will alter user experience. I'm taking notes on how this will change interactions with the system, and how the introduction of ads may alter the perception of the system. I will need to consider the system's business model. I'm focusing on the ethical ramifications and potential for manipulation. I am also working on framing the user as either a commodity or a customer.

info

Raw JSON: 2026-01-21-our-approach-to-advertising-and-expandin-cda-spicy-lite-cbgsaz.json Framework: CDA-Spicy-Lite v1.0 Generated: 2026-01-21T10:37:53.834Z

Discourse Depot © 2025 by TD is licensed under CC BY-NC-SA 4.0