Google May Start Reading Your Gmail and Photos to Target Ads
Google quietly updated its privacy policy to say that data from Gmail and Google Photos could be used to power 'personal intelligence' features and target advertising. The company had previously promised not to scan Gmail content for ad purposes after dropping that practice in 2017.
The keynote reveals 'personal intelligence' as a warm new capability; the mechanism turns out to be the thing they promised to stop doing in 2017.
New name, old inbox.
X algorithm factor 5
Relies on the reader already knowing the Gmail-scanning backstory; without that it's just a shrug of a punchline, and the date error gives skeptics an easy 'well actually' quote-tweet.
Conceptual accuracy 5
The real story says Google dropped Gmail ad-scanning in 2017, but the punchline says '2016' — a factual slip that undercuts the joke's credibility even though the overall gist (old practice, new branding) lands.
Humor dry 6 · nerdy 3 · zinger 5 · pithy 7
Dry setup works ('It knows you now. Deeply.') but the payoff is a flat date-drop rather than a sharp mechanism gag, so it reads clever-adjacent rather than quotable; little technical specificity to reward nerds.
Visual clarity 5
Panel 1 is visually busy with founder squeezed beside an oversized board; Norm's raised-fist gesture in panel 2 doesn't match his questioning line; balloon tails floating above heads rather than pointing cleanly also muddy who's reacting to what.
The dev implements the policy change and discovers it is literally a revert of the 2017 commit that removed Gmail ad scanning.
The diff was small. The promise was smaller.
X algorithm factor 7
The 'git revert 2017' framing is a clean, quotable hook that tech Twitter will immediately get and want to screenshot alongside the news; the closing line is shareable on its own without needing the image.
Conceptual accuracy 7
Nails the actual mechanism (Google reversing its 2017 no-ad-scanning promise) via a sharp dev-culture metaphor, though it never explicitly mentions Gmail/Photos or ads, so a reader unfamiliar with the story might miss the specific target.
Humor dry 8 · nerdy 8 · zinger 7 · pithy 8
The git-revert conceit is a genuinely nerdy, well-built joke that rewards knowing version control, delivered with dry understatement and a tight final zinger that lands as a standalone line.
Visual clarity 8
Panels are clean, boards are legible, balloons sit clearly over the right speaker, and the visual beats (ticket drop, alarmed closeup, deadpan solo) track the emotional arc well despite minimal staging.
A normal person meets personal intelligence at home: the ads have read his email and are being very thoughtful about it.
The condolences are personalised.
X algorithm factor 7
The punchline 'Sorry about your dad, urns are 20% off' is a clean, quotable gut-punch that works standalone without needing the Gmail context explained—people will quote-tweet it as their own creepy-ad story.
Conceptual accuracy 6
Captures the emotional betrayal of the broken 2017 promise well, but the specific mechanism (Gmail/Photos content feeding ad targeting) is only implied, not shown; a reader unfamiliar with the story gets the vibe but not the 'Google reads your email' specifics.
Humor dry 7 · nerdy 3 · zinger 8 · pithy 8
The zinger is sharp and the understatement of grief-meets-upsell is dry and effective, but there's zero technical/nerdy angle on how personal intelligence actually works, making it more generic 'creepy ads' humor than a tech-insider joke.
Visual clarity 6
The 'LLM' label on panel 3's character is a jarring, unexplained left-field choice—it's not a chatbot interaction, it's an ad notification, so slapping an LLM sprite in physically next to the user muddies who's 'speaking' and why an LLM is discussing urn discounts.
In the policy review, the reversal is framed as a clarification, and the intern asks which direction it clarifies in.
Nothing changed except what happens.
X algorithm factor 6
The closing line 'Promises expire. Policies update.' is tight enough to be quote-tweeted on its own, but the strip never visually signals Gmail/Photos/ads, so it reads as generic corporate-doublespeak content rather than a scroll-stopping callout of this specific story.
Conceptual accuracy 5
It nails the 'reversal reframed as clarification' angle but omits the actual subject matter entirely — no mention of email scanning, photos, or ads, so a reader unfamiliar with the story could not reconstruct what happened, only that some vague policy changed.
Humor dry 7 · nerdy 3 · zinger 7 · pithy 8
The dry corporate non-answer and the final zinger work well as a satire of PR-speak, but there's zero technical specificity (no nerdy payoff) and the humor rides entirely on language rather than visual gag, so it's a one-note joke.
Visual clarity 5
Clean three-panel setup with legible board text and distinct speaker labels, but panels 2 and 3 are just floating heads with no scene continuity, and the 'exec' is unlabeled while 'intern' badge does double duty as the only visual anchor.
Down in the datacenter the model has already read every inbox and has formed opinions, which it reports as ad inventory.
It read everything. It has notes.
X algorithm factor 7
'Loneliness converts' is a sharp, quotable closer that could travel as a standalone caption on ad-tech cynicism, and the creepy-intimacy angle invites people to share their own 'it read my inbox and knows' jokes.
Conceptual accuracy 6
It nails the chilling implication of Gmail-scanning-for-ads but never gestures at the actual news hook (the quiet policy reversal, the 2017 promise) — someone unfamiliar with the story gets 'AI reads email for ads' but not the specific broken-promise angle.
Humor dry 7 · nerdy 5 · zinger 8 · pithy 8
The zinger is genuinely good and dry, landing on the dehumanizing logic of ad-tech in two words; it's pithy and quietly escalates panel to panel, but it's more emotionally cutting than technically nerdy — there's no joke about embeddings, scanning, or policy language.
Visual clarity 7
Clean three-panel staging with identifiable LLM-labeled figure, readable balloons, and legible beats, though the middle panel's tight close-up slightly breaks the datacenter setting established elsewhere.