The Exponential Rise of AI: A Conversation on Speed, Accessibility, and the Human Edge

By Biz2Biz Consultants (with AI collaborator) March 15, 2026

Alright, let me weave in a little some breathing room—nothing overwritten, just clearer lanes so anyone reading this doesn’t feel like they’re sprinting through a tech manual. Think of it as turning up the lights a bit, letting the ideas settle without rushing past them. We’ll start with the timeline, but slower:

Two years back, March 2024, AI felt like a clever sidekick. GPT-4 could handle thirty minutes of focused work—say, writing a short script or summarizing a report—before it started rambling or forgetting what you asked. That’s all. No long-term memory, no real independence. Labs were racing, but it was still a niche thing—maybe ten big players, compute at ten-to-the-twenty-fifth FLOPs. You’d chat with it, get help coding, but it wasn’t thinking. More like a super-fast search engine with personality.

Jump to March 2025. Suddenly it’s holding an hour straight: train a basic model, debug half a program without derailing. Compute doubles—ten-to-the-twenty-sixth—and companies jump in hard. Gartner says seventy-eight percent are using it now, up from fifty-five. Claude 3.5 starts stringing thoughts together; GPT-5 previews tease what’s coming. It’s not just faster—it’s longer. Like going from a sprint to a jog. People start noticing: hey, this isn’t a toy anymore.

September 2025? That’s when the floor drops. Eight-hour stretches—full-day coding, research loops, agents that grab tools and run. Siri gets rebuilt, coding copilots hit every dev’s laptop. Compute explodes—dozens of models cross that 10²⁶ line. It’s no longer “help me”; it’s “do this while I sleep.”

And now, March 2026—we’re at two-day projects. Claude 4.5 Opus nails forty-eight hours with over half the time it actually works. Training’s north of 10²⁷ FLOPs. Global spend? Two-point-five trillion. IMF says AI’s already adding one-point-three percent to yearly GDP. Every six, seven months—boom, double again. Not magic. Just relentless math.

Here’s where it gets real: most folks aren’t tracking this. Pew’s got ninety-plus percent saying “yeah, I know AI”—but ask what it means? Crickets. Teens use it daily—sixty-four percent—but they’re not dissecting it. Verasight poll: sixty-four percent touched it last month, half weekly. Yet fifty-six percent are nervous, sixty-three percent scared we’ll lose the reins. Half the country’s more worried than thrilled. Younger ones lean on it for feelings—thirty-four percent say “emotional support”—but they’re also the most stressed about everything collapsing. It’s like they’re riding a rocket they don’t understand. And yeah, even the builders are lost. X threads full of devs going, “I spent three days mastering Claude 4.5’s new chain-of-thought—then a patch wiped it.” Updates every fourteen days? That’s not evolution; that’s churn.

For regular people? Grok spits perfect code today, blanks tomorrow unless you subscribe. No gentle onboarding—no “here’s how to start” like the old Nokia days. It’s firehose. Overwhelm isn’t a bug; it’s the feature. I keep coming back to the dot-com crash—same rhythm. Bay Area kid, I watched it: hype, billions burned, crash, then the winners—Google, Amazon—rose quiet. Today’s AI? Same hype, bigger money—trillions—and it’s not just ads; it’s rewriting how we reason.

Right now it’s messy: clunky prompts, hallucinations, constant resets. But after the bubble? Voice-first, no-tech-speak helpers. That’s the iPhone shift. Until then, learn the guts—prompting tricks, spotting lies, keeping context. Those who do? They’ll surf the wave. Everyone else drowns in noise. Kids treat it like wizardry—essays in seconds, art from text, apps built overnight. Entry gigs? Junior dev, copywriter, designer—AI does ninety percent cheaper. By 2030, those ladders vanish.

New rule: don’t compete with the machine. Steer it. AI’s brilliant at patterns, terrible at invention—can’t feel heartbreak, can’t write a poem that aches because it never ached. Humans? We leap. We lie to ourselves and still find truth. To the youth need “AI skepticism” now—schools need to teach it, question every output, remix, call out BS. That’s the edge.

Customer service? Brutal. I remember clerks who looked you in the eye, fixed your mess, maybe cracked a joke. Now, it’s bots: “I’m sorry for your inconvenience.” No warmth. Kids grew up with Siri—they don’t miss it. But MIT says eighty-three percent still want humans for the real stuff: refunds, breakups, rage. Not data—connection.

AI clears the boring, so maybe real talk gets room. Jobs that survive? Anything messy: therapy, sales, raising kids. People smell fake a mile off. When a teen’s world cracks and the bot just echoes “I’m here,” no hug—watch them turn back to what’s real. Human touch? It’ll be luxury. Like handwritten letters after email. So yeah, AI’s flooding in—calendars, bills, chit-chat—so what’s left is us. The raw bits: eye contact, silence that means something, “I screwed up” with weight. Fun? Absolutely. Real? Never. Kids who master both—tech plus heart—won’t fight for scraps. They’ll shine. Not because they’re faster. Because they’re alive. The only thing that never gets patched out.

But here’s the quiet truth no one says out loud: we might already be too late to rewind. The kids who grew up swiping before they spoke—whose first “sorry” came from a robot, whose first fight got resolved by a thread—don’t know what real warmth feels like. Not because they’re broken, but because they never got the demo. No grumpy cashier who still smiled. No barista who remembered your order without asking. Just clean, cold, instant.

And yet… the ones who do remember? The outliers who still look up when you talk, who pause before they answer, who say “I get it” and actually mean it—they’ll be rare. Like vinyl in a streaming world. Not because they’re better at tech—they won’t be—but because they’re better at being human.

In a decade, when every checkout line is voice-only and every complaint gets a scripted “we value your feedback,” the people who can still do eye-contact, still laugh off a bad day, still fix what a bot can’t… they’ll be craved. Not hired. Craved. Like finding someone who still writes thank-you notes.

So yeah—AI will eat the routine. But the real job? Staying human. And if we don’t teach it now, the next generation won’t even know what they’re missing. Until they do. And then they’ll pay anything for it.