Gen AI Leaders, It’s Time to Learn the Lessons from IoT Hype
A decade ago, the Internet of Things (IoT) was hailed as the next big thing. I should know; I was there as the founder of the Internet of Things Consortium.
After years of industry conferences heralding the “Year of Mobile,” talk in 2012 and 2013 was about how IoT was soon going to be everywhere. The space was already attracting a surge of startups, venture capital investments, and a lot of talk.
There was wide speculation that wearables and connected home would change media consumption, shopping, and the still incipient rise of social media. By 2015, when the Apple Watch debuted in 2015, the 4.2 million sales in the second fiscal quarter of that year led many to believe that the iPhone would be soon supplanted by an all-in-one timepiece.
While many households started getting “smart” thermostats, security systems, wifi-enabled appliances, the boom never really materialized. Even in the start of the IoT hype cycle, CES attendees tended to mock the smart fridge that would order Fresh Direct when it was low on eggs.
The idea was that consumers are passive, but they’re not that passive. Ordering groceries doesn’t require that much assistance.
These days, generative artificial intelligence is inspiring even greater promises. It will handle all the mundane tasks and answer all the questions everyone from consumers to creators might have. It’ll write the news and it’ll even choose your vacation destination and handle the ticket purchase, while simultaneously buying clothes for the trip.
Again, consumers are passive, but they’re not that passive. The most realized promises of technology are the ones that actually do make tasks easier and take the friction out of our busy days. It’s why a “smart car” like Tesla paved the way for electric cars — it made “smart” cool and sleek, in addition to being more fuel efficient.
Hardware Is Hard
Let’s stop there for a second. Hardware is hard: it’s costly to manufacture, it’s costly to distribute. It’s costly to update. An initial sales boost for the “shiny new toy” rarely is able to be maintained.
The scenario happens so often, it hardly bears repeating: All it takes is for one hit product to make an impact after years of iterating and struggle. Then, within months — maybe even weeks these days — others quickly jump into that market and make similar ones at a cheaper price point.
On top of that challenge, hardware is constrained by global supply chains, which can shift at any moment and turn fortune into failure.
That was one of the primary problems plaguing the rollout and growth of many IoT products.
Software, on the other hand, is more easily protected, easily licensed, and easily upgraded. Last year’s cool hardware design is more quickly dated these days, whereas last year’s software gets upgraded while the customer sleeps.
Software, by its nature, is eminently scalable. Hardware gets sold for ironic pleasure on eBay. Software is the overwhelming difference and advantage that generative AI has in comparison to the IoT era. But that doesn’t mean it’s immune from other issues that made things tough for IoT startups.
Grappling With Hype Waves
Looking back at the IoT’s early days, FirstMark Managing Director Matt Turck, a keen tech investor at the dawn of that era, posted his frustration on LinkedIn with how the concept lagged as a cohort of entrepreneurial ventures.
While Turck conceded some points in the comments, he expressed dismay that only one single independent public company, automotive fleet tracking platform Samsara, emerged from the IoT frenzy. The majority of the IoT market's value ended up in the hands of established giants like Honeywell, Comcast, and Big Tech.
As the gen AI industry grapples with its own enormous hype wave, it’s worth taking a practical look at the lessons investors and startups can learn from the IoT phenomenon, and apply them in some form to the future of AI.
Still, don’t count IoT out. While consumers may not have adopted wearables en masse, and few IoT startups became household names, connected device software and hardware has made its way into the b2b space.
According to McKinsey & Co., the potential economic value of IoT is still large and growing. By 2030, the consultancy estimates that it could enable $5.5 trillion to $12.6 trillion in value globally, including the value captured by consumers and customers of IoT products and services.
The breakthroughs and value are primarily focused in healthcare, transportation, and smart cities initiatives. It may not inspire consumers’ desire, but IoT — and Industrial use cases are burgeoning.
In part, the rise of IoT among business leaders reflects the appeal of the metaverse — aka, hybrid digital/physical experiences, or “spatial computing” as Apple is defining it — in less attention-getting circles.
A study from EY and commissioned by Nokia this summer noted that virtually all of its 860 business leaders who responded to a survey on “Industrial IoT” expressed an abiding faith in the revenue-generating capabilities of metaverse applications.
“Nearly all (96%) respondents see how, by mixing physical and virtual use cases, the metaverse brings additional innovative capabilities that will allow them to accelerate the deployment, adoption and monetization of Industry 4.0 for their business,” the study found.
Of course, that utilitarian focus often gets overlooked when the main focus is on, “What does it mean for the tech giants?”
Influence of Larger Companies
In both the IoT and AI industries, the influence of major tech companies, including Amazon, Apple, Google and Microsoft, cannot be ignored. These giants wield significant power and resources, which can dominate the market and overshadow smaller players. Entrepreneurs and investors in the AI space should be mindful of the influence of these tech behemoths.
Regulatory challenges are another area to expect some similarity, though gen AI events are moving much, much faster than anything except the emergence of the dotcom era almost 30 years ago.
IoT faced regulatory and privacy challenges in the U.S., which impacted its growth. Privacy and security concerns, combined with regulatory hurdles, can slow down industry progress. The AI industry should appear ready for those government incursions.
What is it good for?
One of the pitfalls of IoT was the lack of clear use cases, particularly in the business-to-consumer (B2C) sector. For AI to succeed, it must provide tangible value and solutions to end-users. Identifying and addressing specific use cases will be crucial for AI entrepreneurs.
For example, IoT struggled to align with the objectives of Fortune 1000 companies. Similarly, AI startups need to ensure their solutions align with the strategic goals of larger enterprises to achieve meaningful adoption and success.
Interoperability issues, meanwhile, have dogged IoT and the connected home to this day.
Gen AI faces interoperability critiques as well. Ensuring that different AI systems and technologies can seamlessly work together will be vital for the industry's growth.
Founder exodus was another problem for IoT companies.
But even more than the brain drain at those pioneers, IoT faced dilemmas between building new infrastructure and digitizing existing systems. AI entrepreneurs must carefully assess whether to build from scratch or integrate with existing infrastructures.
Media misperceptions is another part of the IoT coverage from 10 years ago that persists. IoT companies, in particular, often struggled to gain mainstream press attention, impacting its visibility to consumers. The AI industry needs to do some more work on effective messaging and storytelling to capture the attention of both financial and mainstream audiences.
Granted, there are differences between IoT and gen AI that are too numerous to mention. But as the future of gen AI plays a small, yet substantive role in all kinds of legal and commercial scenes. For example, take this week’s US Department of Justice antitrust case against Google’s search dominance, entrepreneurs and investors in the space should be wary of past hype.
As I told The Drum’s Kendra Barnett, “The trial is further proof that the AI future is already likely to be organized and dominated by the existing giants as opposed to entrepreneurs seeking to move technology’s power centers to new ground under new terms. This momentum ‘should be of particular concern to advertisers and brands worried about the next chapter of the media landscape.’
It's been an interesting journey
The Internet of Things also focused on one type of connection, leading to security concerns, especially where other solutions using embedded systems or private connections were available. There is also a focus on the Things, and not on the interactions and exchanges of all parties within the ecosystems where these Things must exist. This leads to unsustainable and unattractive Cloud architectures and subscription business models.