Generative AI

Early Adoption: Imagining What’s Possible with Generative AI and Apple Vision Pro

At WillowTree, we like to ask questions, dig deep, and poke holes in any new technology — and then find better ways to repair them.

This inquisitive approach lends itself well to early adoption: we want to understand how emerging tools and tech work — like the Apple Vision Pro, for example — to assess their potential and risks.

We’ve found this strategy especially helpful and relevant in the generative AI space, where our teams encounter questions about what an AI-driven world might look like, such as:

  • How can we use AI to offload time-consuming tasks?
  • What efficiencies can we expect to gain?
  • What are the risks to avoid and opportunities to pursue?

Our predictions may be flawed.

What’s more important? Focusing on what’s possible.

Let’s look to the past for inspiration.

New York Times technology columnist Erik Sandberg-Diment exemplifies how best to think about commercial technology in its early phases.

In the 1980s, Erik was a renowned critic of new-to-market technologies. He was known in Silicon Valley for his weekly column filled with colorful product reviews and predictions on which new technologies would flop first. More than anything else, he was a skeptic. Consider this excerpt from a column published on December 8, 1985:

“The real future of the [blank] will remain in the specialized niche markets. Because no matter how inexpensive the machines become, and no matter how sophisticated their software, I still can't imagine the average user taking one along when going fishing.”

Before you read the column, which device do you think Erik was referring to? Hint: it’s one of these four:

  • CD player
  • Disposable camera
  • Laptop
  • Portable gaming device

The 1980s: a booming time for tech innovation

Erik was writing about the launch of the portable laptop computer. (Was that your guess?)

When Erik wrote this article almost four decades ago, computer chips were becoming smaller, shrinking from 12,000 nanometers in 1970 to 3,500 nanometers in 1980. (For reference, today, these parts measure about 10 nanometers in diameter, smaller than the average virus.)

This remarkable innovation meant companies could create more portable technologies and devices than in years past, and laptops became available for everyday consumers to purchase and use. Still, the average cost was $7,000 (almost $20,000 in today’s dollars) for an 18-pound “portable” device with minimal software.

Again, I can’t help but notice the resemblance between early laptop critiques and initial skepticism toward the Vision Pro. High-end price tags. Bulky, unfamiliar tech. Consumer preference. Sounds familiar, right?

Back then, Erik wasn’t buying it.

His job was to investigate and communicate the promises and perils of new technologies. And, in this case, he was confident the general public wouldn’t find value in an expensive, bulky device with limited features.

“For the most part, the portable computer is a dream machine for the few,” Erik wrote.

But here’s the key point: though Erik initially saw the laptop’s size and cost as hurdles for the average consumer, he nonetheless imagined the possibility of a more accessible use case. He proposed that a laptop with limited software could succeed in specialized markets — such as inventory management or sales — where on-the-go tracking and quick communication between field technicians could positively impact a company's bottom line.

That’s what made Erik’s writing so valuable. He could envision the promise of various kinds of technology while simultaneously critiquing the version before him.

Like Erik, many of us at WillowTree are imagining what’s next for AI tools and still constructively critiquing their capabilities today. We envision a future state where AI adoption is more widespread, secure, and accessible. WillowTree growth marketing experts Billy Fischer and Billie Loewen recently discussed various exciting, innovative LLM and AI use cases on our Room for Growth podcast.

But at the same time, we’re taking a methodical approach to assessing the risks and outputs of AI tools. My product research colleagues Jill Stover Heinze and Nima Meghdari recently shared their perspectives on crafting GPT prompts to generate results that are both valuable from a business perspective and ethical from a human perspective.

It’s a relevant approach to understanding new and emerging tools beyond AI. Our President Tobias Dengel sums it up well in his article on spatial computing and Apple’s latest WWDC announcements:

“Any new tech requires thoughtful scrutiny and patience before we can guess at long-term implications, and the initial use cases and adoption rate are not yet clear. That said, as strategists, developers, and designers, we’re excited to explore the new platform.”

Yes, we can carefully consider and critique today’s limitations for immersive tech like Vision Pro. But it’s our responsibility as digital experts to also “dream” of what’s possible. As Tobias discusses, the launch of Vision Pro has prompted us to imagine wide-ranging spatial computing use cases that could revolutionize sectors like healthcare and wellness, media and delivery, and consumer goods and retail.

I think Erik would appreciate all of our teams’ imaginative yet investigative approach to implementing these ever-evolving tools.

Conversations with early chatbots in the ’80s

Even in 1985, Erik similarly evaluated artificial intelligence. He wrote about Symantec's word-processing and database package that was “supposed to allow the user to tell the computer in plain English what to do.” He also experimented with entertainment software called Racter that “permits a computer owner to have a spontaneous, albeit bizarre, conversation with [their] computer.”

Erik explained his confusion with some of Racter’s responses to his questions. He wondered about the specific material and subject matter the program was trained on. And he described AI’s potential as a user’s “guide through the maze of stored knowledge.

His musings resemble today’s conversations about natural language processing (NLP), large language models (LLMs), data privacy, and AI hallucinations. ChatGPT can not only access bottomless reservoirs of knowledge, answering questions about an infinite number of subjects, but it can also create from scratch  — a term paper, perhaps, or a software application.

Sam Altman, OpenAI CEO, has acknowledged that this new technology could result in various positive and negative outcomes. Now is the time to think hard about those possibilities.

Even though Erik wasn’t always right in his predictions, he interrogated his subject matter intensely. He was transparent with his audience about how he reached his conclusions.

Erik performed due diligence when reviewing new technology — investigating tools carefully to understand how they worked and highlighting benefits and risks.

He finds the delicate balance that we technologists have sought for decades, maybe centuries, confidently walking the wire-thin balance beam between moving quickly and moving thoughtfully.

We all benefit when early adopters investigate and critique new tools.

Let’s return to Erik’s example of the laptop.

As of 2022, about 68 percent of US adults aged 18 to 64 owned a laptop, representing more than $21 billion in revenue. Given the popularity of the laptop today, some might say that Erik was wrong to criticize the portable computer when it first emerged.

He might say that, too. “So I was wrong about laptops,” he admitted to the Atlantic in 2016.

“But I turned out to be right,” he added. In a world where smartphones are now dominant, “the laptop is at the bottom of the curve.”

In the prediction business, sometimes you can be correct by accident. But nowadays, Erik has also expressed concern about how personal computing has left us all a little bit more isolated. His predictions could be a function of that skepticism just as much as his optimism.

Regardless of whether those predictions were correct, Erik used his columns to critique technology as it existed at the time while contemplating and preparing for a different kind of future.

This process requires a profound act of imagination, something we try to do daily at WillowTree. We use best practices to investigate, critique, and evaluate tools and technologies as they are available today. But because the world never sits still, we consider their potential for tomorrow.

Once more, Tobias summarizes it well as he considers Vision Pro: “We're excited to explore these new horizons and help shape the future of digital interaction.”

We see risks. We see challenges. But above all? We see opportunities.

Get in touch to learn about our eight-week GenAI Jumpstart program and future-proof your company against asymmetric genAI tech innovation with our Fuel iX enterprise AI platform.

Read the Video Transcript

One email, once a month.

Our latest thinking—delivered.
Thank you! You have been successfully added to our monthly email list.
Oops! Something went wrong while submitting the form.
More content

Let's talk.

Elegant, Performant Digital Products.
Personalized, Automated Marketing.
The Frontiers of Data and Generative AI.