From Spellcheck to Ambient AI: Why the Best Tech Eventually Disappears
The assimilation of spellcheck freed us to focus on ideas instead of i-before-e.
This issue of Social Signals was written to You’ve Been Spiked by Chris Joss.
Hello from the South Shore of Lake Superior (aka the northernmost part of Wisconsin), where I rented an artist’s studio on a cliff overlooking the largest freshwater lake in the world for a week away from roofing solicitors, good wifi, and well.. noise.
I read books and drank good coffee and spent time being quiet and thinking. As a family we went for hikes, threw rocks in the lake, played board games, watched movies, and learned about lamprey (gross!). Here’s a little video recap.
Books:
We only have a few weeks before our oldest leaves for college, and it was an intentional trip to spend time making memories together and enjoying each other. And it was a massive success.
Hoping you each get a few moments to savor the summer and your people. It’s always a worthy investment. Anyway, let’s get into it! -Greg
From Spellcheck to AI: How the Next Tech Wave Will Become Invisible
I was 10, wearing my new LA Gear Regulator Pump shoes, and I still remember the adrenaline rush of spelling “chrysanthemum” correctly to clinch my class spelling-bee crown. (Peak life achievement? Not a chance.)
Fast-forward to 2025, and not one of us fires up www.spellcheck.com to double-check if the word “refrigerator” had a D in it. It doesn’t. I think. I honestly can’t remember, but Substack would’ve told me if it was wrong. Because…
Spellcheck just happens. Quietly, in real time, in every app.
But don’t forget that spellcheck was hated by many when it came out. Teachers were worried kids wouldn’t learn the basics of English if they could crutch on tools. I even have a family member who told me that “spellcheck is a sin” and “people should know how to spell words.” True story.
And yet spellcheck came into every app, and we assimilated. We don’t care what dictionary it was trained on or whether it consumes environmental resources. It’s everywhere, and we all have more or less the same access to the notion of spellcheck, even if we’re using different versions and in different ways.
We forgot to be mad about spellcheck, and now we all just use it.
Spellcheck is the quintessential example of technology going from headline feature to invisible infrastructure.
Today generative AI is racing down that same path. Let’s unpack what that means for marketers, creators, and anyone with a keyboard.
A Quick Spellcheck Time-Hop
1960s–70s: Mainframe programs like Stanford’s “SPELL” run in batch mode. You feed the beast your file, grab coffee, then sift through a printout of suggested fixes. Maybe it was better to just know how to spell words?
1980s: WordPerfect and WordStar bundle spellcheck, but you still hit F2 to launch it. Many of us remember this as the calculator app of its day. I used it ALL THE TIME.
1990s–2000s: Red squiggles arrive! Microsoft Word starts flagging typos as you type. Clippy applauds. Damn it, Clippy!
2010s: Apple bakes system-wide autocorrect into macOS and iOS. Grammarly layers on cloud-based grammar suggestions. There is suddenly no separate workflow. Friction evaporates.
2020s: Spellcheck is everywhere, all at once: your phone, your car console, your smartwatch, your email, your software, your MS Teams. Companies force Grammarly onto your work computer to cover literally everything. Spellcheck is so ambient that people forget it was ever “a feature.” It’s a utility.
Fun Glitches on the Road to Grammatical Perfection
The journey of spellcheck wasn’t without its quirks. Maybe you’ve heard of these?
The “Cupertino Effect” - Early autocorrect swapped “co-operation” with “Cupertino,” sprinkling the Apple headquarters name into many documentions unintentionally (including a UN memo, whoops!).
The Ghost Word “Dord” - We can go back to those printed dictionaries and find glithces, too. Webster’s accidentally printed a non-word in 1934. It’s not a word. It’s a dord. See? Even dead-tree spellcheck struggled with data hygiene.
“Candidate for a Pullet Surprise” - This is an early meme that was passed around in emails in the 90s (remember when memes were shared via email) and poked fun at over reliance on spellcheck by rewriting a famous poem in a way that would pass spellcheck but was full of incorrect words.
I have a spelling checker,
It came with my PC.
It plane lee marks four my revue
Miss steaks aye can knot sea.
AI Is on the Same Conveyor Belt as Spellcheck
Here’s where it gets interesting: AI is following the exact same path as spellcheck. Right now many teams still “launch” AI: they open ChatGPT, fire up Midjourney, or click a Copilot button. That is 1985 all over again.
Right now, we still treat AI like the calculator app on our phones. Like something we choose to open when we need it. But that’s about to change.
Just as spellcheck became a seamless, always-on feature, AI is being embedded into everything.
Microsoft Copilot is fused into Windows for code snippets, slide outlines, or Photoshop-level image edits.
Google Workspace with Gemini autocompletes paragraphs, charts data, and your Gmail replies to incoming messages with suggested responses and in your tone of voice.
Adobe Firefly turns one-line prompts into polished design elements, directly inside Photoshop and Premiere.
GitHub Copilot pipes suggestions inline while you type, accepted with a tap.
ChatGPT launched “Record Mode” that will record and transcribe meetings even if your meeting software doesn’t provide summaries and action items.
Every phone launch keynote now touts on-device AI that summarizes notifications or rewrites your one-word text so Mom knows you love her.
The ignition sequence is fading away. AI is slipping into the OS layer. Not just the OS of our devices, but the OS of our lives. Before long you will simply expect your tools to anticipate intent, just as you expect a red squiggle to appear under “teh.”
Leveling the Field (The Scary-Cool Part)
This is the part where skeptics ask: Doesn’t that mean everyone will have access to the AI tools? Won’t it stop being a competitive edge? Exactly. And that’s the point.
It’s the same thing people said when spellcheck hit desktops: “Now every memo will be perfect.” Nope. Perfect spelling did not create great prose. The launch of Excel did not mean less accountants; it meant we could do greater reporting and math!
Likewise, ubiquitous AI will not replace the need to know how to do things. At least not all of the things.
It will, however, vaporize rote tasks (ahem, drafting, formatting, summarizing, synthesizing, heck even AI agents doing the writing and sending for you), so the value shifts to taste, judgment, and bold ideas. In other words, creativity becomes more visible once the mechanical stuff is handled by the machine.
I welcome it.
AI isn’t here to replace the human spark. It’s here to amplify it.
When Tech Becomes Invisible, Culture Shifts
Invisible tech rewires expectations:
Readers skim faster because auto-summaries exist.
Audiences demand personalized everything because AI easily tailors it.
Teams sprint from concept to prototype in an afternoon, and strategy cycles shorten.
The winners will be the folks who treat AI not as a magic trick but as running water: always on tap, rarely celebrated, utterly indispensable.
Treat it as if it will always be here and be integrated into everything you do, because it probably will.
So what will you do with the brain-space you just got back?
I don’t do spelling bees anymore. And I honestly cannot remember whether “refrigerator” has a “D,” and it hasn’t mattered for decades.
I’ve given that part of my brain to other things and let spellcheck take over. Does that make me less human, or more human?
The disappearance of spellcheck freed us to focus on ideas instead of i-before-e.
The spread of AI will do the very same thing, only at warp speed. The real flex now is not knowing every step of the process, it is knowing where you can leapfrog the process to deliver something that surprises and delights.
Every time a technology fades into the wallpaper, it hands us a chunk of creative bandwidth. So…
Spend that bandwidth recklessly. Draft the risky headline. Prototype the weird concept. Ask the question that makes the room go quiet, then makes the work get better.
Because in a world where everyone has the same invisible copilot, the only true differentiator left is you: your taste, your curiosity, your willingness to chase an idea past obvious.
So lean in. Automate the ordinary. Double-down on the extraordinary.
And if you ever do need to spell “chrysanthemum” again, the machines have your back. Use that freed-up mental RAM to make something the machines could never imagine.
Your move, humans. -Greg
📢 See Greg Speak
Digital Summit Minneapolis: Who Knew the Best Brand Advocates Already Work Here?, Minneapolis, MN - August 6-7
University of St. Thomas: Marketing, Insights, and Communications Team Onsite, St. Paul, MN - August 12
💥 JUST ANNOUNCED ➡️ Minnesota Blogger Conference: The State of Blogging in 2025: The Blog Is Dead, Long Live the Blog, Minneapolis, MN - September 13 (GET TICKETS)
PRSA Iowa's OctoPRfest Conference: Reboot Required: Rewiring Strategy for Today’s Cyborg Audiences, Des Moines, IA - October 9
Interested in Greg speaking at your event? Hit me up.
📊 Chart of the Week:
🚨 Check Out FINN Partners’ New Crisis Simulation Tool
This week FINN Partners launched CANARY FOR CRISIS, an AI-powered training platform designed to help brands respond to real-time reputational threats fueled by misinformation, trolls, and algorithm-driven chaos.
Built by my longtime colleague and friend and crisis veteran David Krejci, the tool simulates high-pressure scenarios that mirror today’s media ecosystem (aka where traditional and social channels collide in minutes) and factors in his Media Forensics approach that digs into the bots and bad actors who are often shaping misinformation to create or shape a crisis.
If you’ve never been in one of Krejci’s drills, what happens is that participants are immersed in realistic drills involving fake news, deepfakes, and hostile reporters, all in a safe, private environment. And then you get to practice how you would respond on social, in PR, and across your communications organization.
It’s crisis comms scrimmage for the algorithm age. Test your playbooks, people!
Read more about the launch this week in PR Week, on our website, and in this wonderfully GEO-optimized press release.
‼️ Throwback Google Alert of the Week
I love this quote: “If you look at your Facebook and Twitter feeds, our friends lead amazing lives… that’s not real life, that’s success tehater, that’s us perfectly orchestrating that.” A good reminder here even 12 years later.
⚡️ Social Signals
Last month I told you about a Walmart in Tennessee that opened a rentable podcast and video production studio as a signal of the mainstreaming of creator culture (reminder: paid subscribers to Social Signals have access to the full archives). This week here in the Twin Cities we saw the opening of Minnesota Podcast & Content Studio in a suburban strip mall by former TV journalist Dawn Stevens, adding to the growing number of affordable, rentable studios in the metro (I count more than 20 via Google Maps). I don’t doubt the demand for more professional consulting and setup for the growing creator economy, and the move to open a standalone studio between an Ace hardware store and the Pizza Ranch in the exurbs is a signal worth watching.
Mercedes-Benz is integrating Microsoft Teams video calls into your car’s dashboard system, allowing drivers to appear on camera during meetings while in motion. Finally? The feature, debuting in the new CLA, blocks screen sharing but still streams your face. So yes, your coworkers can see you driving. Ummm…
Roblox is launching a new “Trusted Connections” feature that lets verified teens chat with friends. To unlock it, users must submit a video selfie, which AI analyzes to estimate their age. It’s part of a broader push to keep minors safe on-platform and reduce migration to riskier spaces like Discord. It’s also part of the signal around proving you are human online (we’ve been talking a lot about that on Social Signals this past year). Critics say this opt-in system puts too much responsibility on kids and parents, with privacy concerns around biometric data and potential loopholes. I personally believe we’re getting closer to needing iris scans. Good thing I already did mine.
3D artist Austin Beaulier built a sandcastle, scanned it with his phone, turned it into a game level he could run around inside, and then 3D printed it into something that will last forever. This technology is readily in our hands! Also, he calls it his ScannedCastle. Ha!
ChatGPT Agent launched. Another massive signal toward the future of how we’ll work and live. Watch this video.
YouTube is phasing out its classic Trending page and “Trending Now” list, replacing them with category-specific charts like Trending Music Videos and Top Podcast Shows. The signal here reflects changing viewer habits and the rise of micro-trends, with discovery now driven more by personalized recommendations than a one-size-fits-all list.
Webcomic of the Week:
Dr. GPT of the Week: Wired wrote about patients turning to ChatGPT for medical insights, something I’ve personally done with great success.
Share This With Your Friends Who Think Their Phone Listens To Them: A new study reveals how thousands of Android apps quietly track users’ indoor locations using Bluetooth and Wi-Fi, including your location, habits, and even who you’re with. It’s not listening, but you’ve given your apps permission to know enough it may as well be.
Tool of the Week: So given the above, if it feels like the ads chasing you across the internet know you a little too well, it’s because they do. So here’s a tool that opens 100 tabs of pure madness to fool trackers into thinking you're someone else. Let me know your results!
App of the Week: Delta Camera is a Game Boy Camera-style app for your iPhone that’s coming later this year. Sign me up! I’d love to see a brand solely use this for a full month.
Study of the Week to Delve Into: A new study suggests humans are starting to adopt the language patterns of ChatGPT, with words like “delve” and “underscore” showing up more in speech. Researchers say AI is subtly reshaping how we talk, not just how we write.
Word of the Week: Corecore. It’s an aesthetic of feeling.
Thread of the Week: unpopular opinion: Ikea furniture is extremely easy to put together and everyone who says it's incredibly confusing must be kinda dumb
Instagram of the Week: Composer and artist Lucio Arese has developed a systematic data visualization model for birds songs, and they are GORGEOUS.
Important Question of the Week: How would a worm type?
Keep going! 🚀✨
Greg