Smartphones, cognitive decline, and how to push back

Executive overview

Test scores measuring reasoning and problem-solving have declined since the early 2010s — exactly when smartphones became ubiquitous. The mechanism is a cognitive death spiral: attention-economy apps rewire the brain for fast stimuli, making it harder to concentrate and eliminating the activities (like reading) that build intelligence over time.

The fix mirrors how society responded to sedentary office work in the 1960s: we didn't abolish offices, we added deliberate exercise. Today, deliberate cognitive exercise — reading, reflection walks, hard hobbies — is what replaces what we used to get for free.

You don't have to get dumber; you have to treat cognitive fitness like physical fitness.

The data on declining intelligence

• PISA scores (science, reading, math) peak around 2012 and decline sharply ever since.
• Adult literacy tests show a similar drop from the same inflection point.
• Self-reported difficulty concentrating and trouble learning new things both shoot up after 2012.
• Leisure reading among US teenagers collapses at the same moment — "almost every day" readers drop sharply, "hardly ever" readers rise sharply.
• The timing matches the global ubiquity of smartphones, not any earlier technology shift.

The cognitive death spiral

• Pre-smartphone platforms aimed to be useful; post-smartphone platforms aim to be addictive.
• Attention-economy apps offer faster, more desirable stimuli than almost any other activity.
• Reward circuits learn: phone nearby → dopamine hit → reach for phone.
• The phone is always present, so the reward circuit fires constantly — unlike a donut cart that only appears at 4pm.
• Two simultaneous harms: (1) harder to apply existing intelligence; (2) less time on activities that would build more intelligence.
• The spiral compounds: less concentration ability → less reading and reflection → weaker concentration → repeat.

Practical cognitive exercise routine

• Read books — the single best cognitive calisthenic; even easier to read without your phone in arm's reach.
• Use the non-constant companion model: plug the phone in the kitchen; go to it when needed, don't carry it room to room.
• Avoid stimuli stacking — consuming multiple streams simultaneously (phone + TV + second screen); it trains the brain to never sustain a single focus.
• Reflection walks: walk without a device and work through a problem; pull wandering attention back repeatedly.
• Pursue a hard hobby requiring sustained concentration — an instrument, woodworking, a demanding sport — where progress is visible and rewarding.

Q&A highlights

• Decentralized social platforms (Mastodon, Bluesky): The problem isn't who runs the platform or what the moderation rules are — it's the concept of a global conversation platform itself. Hundreds of millions of people on one platform requires such aggressive algorithmic curation that destabilisation is inevitable. Niche communities with their own standards are a better use of the internet.

• High-paying stressful job vs. old lower-paying job: Don't fixate on a specific solution (going back). First clarify the ideal lifestyle — what does the ideal day look, feel, and smell like? Then evaluate options systematically. The new salary may unlock things the old job couldn't; the problem may be reducible with better time management. Focus on the goal, not the grand gesture.

• 40-year-old engineer feeling outpaced by younger colleagues: Younger workers aren't smarter — they're just more recently trained. Systematically learning the current skill each year restores parity. Older workers likely have an advantage in sustained concentration (less stimuli-addicted). Pivot option: move toward managerial roles that reward mature decision-making, which younger workers genuinely can't replicate.

• Production company dream with a year of severance: Complement goal-based thinking with lifestyle-centric planning. Use money as a neutral indicator of value — wait until people pay you, not just encourage you. If revenue validation can happen in six months, go all in. If it'll take years, keep job-searching in parallel.

• Federal job loss during probationary period: Don't cling to a career narrative that's no longer possible. Inventory existing career capital, find a role that rewards it, rebuild, catch your breath, then re-attack. Strategic regrouping is not retreat.

• Maintaining deep work ability during a shallow-work rotation: The rotation is cognitively equivalent to smoking; offset it with active cognitive exercise outside work hours. Read books, practice the non-constant companion phone model, do reflection walks, learn a hard skill. When the rotation ends, the instrument stays well-practiced.

Tech corner: AI's underrated killer app

• Current AI discourse focuses on text generation (writing emails, memos, presentations).
• The bigger near-term productivity gain is text interpretation — using language models as natural-language interfaces to existing software.
• Example: tell a model in plain English what you want to do with a spreadsheet; it translates that into precise macro commands; the spreadsheet does the work.
• This doesn't require massive frontier models — smaller, purpose-trained models can do it, meaning many companies can build their own.
• Large AI labs emphasise text generation because frontier-scale models are their competitive moat; natural-language-to-software translation is attainable by smaller players.
• The real paradigm: language model as translator between human intent and existing specialist tools — not as an oracle that does everything itself.