Toys can tell us a lot about how tech will change our lives

LEGO Mindstorms, PlayStation 2 and Furby all resonate today in their own way

Column Twenty-five years ago this month I published a book called The Playful World that explored a simple idea: that the seeds of the future can be found in the present by considering the dazzling toys we started giving our children at the turn of the millennium.

I considered three of those toys – LEGO's Mindstorms, Sony's PlayStation 2 and the Furby – as I made predictions about the future of the physical world, the virtual world, and artificial intelligence.

Did I get any of it right?

LEGO's Mindstorms may no longer exist as a product, but the language created to program those little blocks – Scratch – has become the de facto first programming language for children around the world.

My biggest miss was a confident prediction that nanoscale manufacturing would be a common technology within a generation.

Early progress in nanotechnology seemed to indicate a future crowded with nanoscale machinery to build new materials, repair our cells, and much more besides. It seemed straightforward until we got stuck on the physics of the nanoscale, a weird mix of classical and quantum worlds that overlap uneasily.

Scientists are still puzzling that out, and although we know far more than we did a quarter century ago, that understanding has yet to translate into much of consequence – with one notable exception: Semiconductor fabrication.

In 2000, I marveled at 1 GHz CPUs. Not long after that, the number of transistors on a CPU soared past 1 billion, then kept growing. The nanoscale manufacturing techniques used to make chips kept pace until around a decade ago, when Intel's wheels fell off. Today, even with TSMC burning much of its profits into next generation processes, we're well behind the doubling-every-16 months we enjoyed for half a century after the formulation of Moore's Law in 1965.

Semiconductor design and manufacturing is now a multi-trillion-dollar industry at the core of the global economy – yet it can't bend physics to its needs.

My predictions for simulations and virtual reality produced mixed results.

At the turn of the millennium, VR had been "dead" for a few years, but tiny and powerful simulators – cunningly disguised as video game consoles – had become familiar objects in many households.

When the smartphone became ubiquitous, it created the potential for VR to go mainstream. Google tried to make that happen with its "Cardboard" VR viewer, a device that kicked off a new wave of enthusiasm for VR that prompted Oculus, HTC, and a dozen other big players to create their own hardware.

The bubble burst at just the moment Meta went all-in on the metaverse. VR might be dead again – Zuck barely mentions it these days – but simulation remains at the core of entertainment, design, and manufacturing.

I missed augmented reality completely, despite its origins in Ivan Sutherland's original work on head-mounted displays. Before Microsoft released HoloLens, augmented reality had been thought a dead end – reinforced by the failure of Google Glass. Today it's the "next big thing" for Meta, Apple and Google, and later this decade might supplant the smartphone as dominant attention thief.

I scored one bullseye in my predictions: Expanding on Sherry Turkel's outstanding work with small children who used Furby, I foresaw a future where those children, once grown, would expect a world constantly listening and constantly responding to them.

A quarter century later, young adults who grew up with interactive toys now engage in continuous "ambient" conversations with ChatGPT and its peers, confiding secrets, seeking advice, getting therapy, and – for both good and ill – being seen. Intelligence now pervades a range of our technical systems, and while it may be far from perfect, it's "good enough" to get us to project all sorts of relational qualities into it that we want it to have.

"The Playful World" drew a line between Furby and ELIZA, the first AI chatbot. Sixty years ago ELIZA elicited our ability to project interiority onto machines. We always want to believe that they think and feel. That they're listening, and they care. This says a lot more about our tendency to anthropomorphize than anything about our machines.

One more thing: My book also examined new forms of our intelligence, centered on a newly released online version of Encyclopedia Britannica. Almost as soon as it launched, Britannica online went down – DDoSed by a global internet thirsting for facts.

That crushing popularity led Britannica to make arguably the most misguided business decision of the early web era: erecting a paywall. The Internet abhors a vacuum and it wasn't long before Wikipedia emerged, filling the factual space better than Britannica – then rendering it obsolete.

Oops.

While the future often defies expectations, it appears people remain stubbornly predictable. ®