Quality Reading,
Without the Noise

Attention interruption and comprehension are directly linked in cognitive psychology research. Studies of context-switching costs — the cognitive overhead of redirecting attention from a primary task to a secondary stimulus — show that even brief attentional deflections reduce performance on tasks requiring sustained comprehension. An advertising banner that activates the attention capture reflex (movement, color contrast, social cues) does exactly this: it taxes working memory that would otherwise be available for integrating the text's argument.

Gloria Mark's research at the University of California, Irvine on interruption and task resumption found that after an interruption, it takes an average of over twenty minutes to return to a complex cognitive task at full engagement depth. Applied to reading: a single advertising distraction during a 2,000-word article may prevent the reader from ever reaching the depth of comprehension the article's later sections require.

Reading dwell time and comprehension quality are positively correlated across studies of digital reading behavior. Readers who spend more time on an article — who read more slowly and deliberately — demonstrate higher scores on comprehension measures, better recall at both short and long intervals, and stronger ability to apply article content to novel questions. Advertising environments, by design, reduce dwell time: visual complexity, cognitive dissonance between ad content and article content, and the interruption cycle all compress the reading session.

Research on reading in educational settings by Naomi Baron, author of Words Onscreen, documents how digital reading environments fragment attention patterns that print reading sustained. Her studies found that students reading on devices with active notification and advertising environments reported significantly lower concentration levels and comprehension confidence than those reading in cleaner environments — even when they could not consciously identify the source of the distraction.

For readers who take their cognitive engagement with long-form content seriously, the ad-free environment is not a luxury but a prerequisite for the kind of reading that actually builds knowledge. Reader Plus is the tier that provides this baseline — and at $1.99 per month, it costs less than a single coffee.

What the AI Companion Does

Article summarization. After you have read a long article, the companion generates a structured summary of the argument's key components — the central claim, the primary evidence, the main counter-arguments, and the conclusion. This is not a replacement for reading: it is a post-reading review tool that activates the retrieval practice effect. Summoning a mental summary before seeing the AI-generated one exercises exactly the active recall that cognitive science identifies as the most effective technique for durable memory encoding.

Vocabulary explanation in context. The companion identifies domain-specific vocabulary in each article and provides contextual explanations — not dictionary definitions, but explanations of how the term functions within the article's argument and within the field more broadly. This approach is grounded in reading comprehension research: vocabulary acquisition is most durable when new words are encountered in rich semantic context rather than in isolation, and explained in terms of concept rather than just definition.

Contextual archive linking. For any article, the companion can surface three to five related pieces from our archive that deepen or challenge the central argument — helping readers build the kind of cross-referenced, multi-perspective understanding that distinguishes genuine knowledge from single-source belief. The algorithm used to generate these links prioritizes complementary perspectives and different disciplines, not just similar topics.

Reading-trajectory insights. Over time, the companion tracks your reading patterns across verticals and provides quarterly insights about your intellectual diet — which domains you are engaging deeply, which you are neglecting, where your reading has been consistent, and where you might benefit from deliberate exploration. These insights are designed to support intentional reading practice, not to create anxiety about performance.

What the AI Companion Does Not Do

The companion does not summarize articles you have not read — the summarization feature requires you to have completed the article before it activates. This design choice is deliberate: the goal is better comprehension through reading, not comprehension shortcuts around reading.

It does not generate new content or complete any reading on your behalf. It does not offer opinions on whether you should agree with an article's argument. Its vocabulary explanations are informational, not evaluative. These constraints reflect a principled position: the companion is a cognitive support tool, not a cognitive replacement.

Understanding this distinction matters for a simple reason. Maryanne Wolf's research on digital reading highlights the risk that reading aids become reading substitutes — that tools designed to help readers engage more deeply end up allowing readers to engage less deeply with a false impression of comprehension. We have designed the companion specifically to avoid this outcome. The reading work itself remains entirely yours.