The AI Hype Cycle Problem
The AI Hype Cycle Problem
AI hype, while a powerful tool for many brands engaged in driving repuation and product interest, may be counterproductive. Misleading…
AI hype, while a powerful tool for many brands engaged in driving repuation and product interest, may be counterproductive. Misleading announcements, weak reporting, and influencer coverage, and a lack of clear value to targeted stakeholders flat-out confuse people.
I have multiple first conversations with Washington, DC area executives every week. Almost half of my initial discussions with new contacts and prospects to date have focused on unraveling vendor announcements and unclear news stories.
Everyone knows and is starting to see AI impacting them. No one quite understands the long-term path. Perhaps that’s because the vendors themselves don’t know. Often, confusion begins at the head.
The AI hype cycle largely overpromises benefits and negative impacts simultaneously. It's no wonder people are confused.
Gartner recently issued its AI Hype Cycle chart (link in caption above), and theorized that the industry is at its hype peak. That of course, means the bubble is about to burst, followed by the trough of disillusionment where people don’t trust technology vendors. That lack of trust is well deserved in my opinion.
Vendor Fueled Vapor?
In our last No Brainer podcast episode, Greg Verdino and I discussed the AI hype cycle. We got into some of the drivers of the hype cycle and offered tips for folks who are looking to discern impactful announcements from spin.
We started by reflecting on past tech hype cycles — the .com era and the rise of web 2.0 — and how to avoid repeating the same mistakes. It became apparent that similar patterns played out across each wave of innovation.
Big companies utilize willing media outlets, industry analysts, and other influencers to amplify their marketing messages as impartial news and analysis. The cycle repeats on a weekly basis, shaping narratives that benefit those companies’ interests.
This is reminiscent of the dot-com bubble when analysts and research firms published overly optimistic predictions that reporters reprinted as fact. The same thing happened in the 2.0 wave, but instead of analysts, you had bloggers amplifying announcements, sometimes trying to scope the media with different perspectives or first-to-market news.
The difference today is the decline of quality journalism and the rise of influencer marketing. With fewer experienced reporters and a fractured media landscape with little incentive to dive into analysis, there’s less scrutiny and context applied to new technology announcements.
Examples of the Hype Cycle at Work
Today, breathless social media hype fueled by media and influencers alike spreads speculative claims faster than ever. This combination makes the public more susceptible to manipulation than even during the .com era.
Recent examples of this include the following:
ChatGPT’s roll-out of “multimodal” AI where you can “prompt” by uploading an image and/or use voice queries to ask it information. This was the second time OpenAI announced this, the first time in the spring with a vague development announcement.
However, it was not immediately available this time, too. Open AI slow-rolled out the features to some GPT Pro and Enterprise users over the past weeks in the U.S. It is slowly making it across the globe. Free users do not have access to these tools. Like almost every recent Open AI announcement, you have to dig deeper to get to the truth.
Anthropic recently announced a major investor, Amazon, officially joining Microsoft and Google with significant investments in LLM technology. Interestingly, most of the media and influencers misreported the story, claiming a full-blown $4B investment in Anthropic.
It was “only” a cool billion with options for up to $4B. Anthropic may choose not to seek investment (in fact, it announced a raise a t a much higher valuation immediately following the announcement) or might need to achieve performance goals to get the rest. Perhaps more interesting was the minority stake, underreported, and the use of Trainium and Inferentia chips to power the AI.
Meta upped its ante in the AI chatbot wars. The Facebook/IG/WhatsApp owned announced its own assistant. It also announced a slew of 28 celebrity AI characters like Charli D’Amelio, Dwyane Wade, Kendall Jenner, MrBeast, Snoop Dogg, and Paris Hilton (eye roll). It’s releasing these features in WhatsApp, Instagram, and Messenger.
Meta AI also announced the ability generate images like Midjourney or OpenAI’s DALL-E via the prompt “/imagine.” It produces compelling high-resolution photos in a few seconds and was likely trained on Instagram (many speculate this, but it has yet to be confirmed). They also previewed AI-enabled eyeglasses.
Again, coverage fails to point out that much of this is promised and not delivered yet. Further, it is interesting how much of it targets consumers within Meta’s universe of applications as fodder to fuel engagement. Media should analyze further… It’s business as usual at Meta, and the questions are: Will people like these new tools, and will competitors respond?
So what should we do about parroted and over-embellished hype disguised as news?
Prescriptive Measures
First, just because someone or some masthead says something about a new AI development does not mean it’s factual. Internet literacy demands a deeper dive, including asking some of the following questions:
Who would use the AI? Consumers? Small businesses? Enterprises? Developers?
Is it available now, and to whom? At what cost?
Is there any analysis of the implications?
Have others reported the news and is their coverage differing, providing a different point of view? Or is it just another aggregator story?
As we ride the AI wave of the 2020s, we should play the long game and apply the lessons of past cycles. Hyped technologies like AI hold tremendous promise and potential business dangers if adapted with sufficient care and wisdom. Businesses need to vet AI technologies before adopting them and ensure that they serve their mission. Let’s also ensure they do so in a secure and ethical manner.
Vendors need to consider their long-term reputation. You can’t control news once you announce it, but be clear about what is getting announced and who it serves. What is the value for the target audience? Is it available now, and if not, when will it become available? Or is your news just a development announcement? Be factual; be clear. If this is hard, consider hiring better product marketers.
Finally, journalists, influencers and analysts who want to become authoritative in the long term must provide the nuanced scrutiny and perspective needed to guide wise adoption. With conscientious reporting and analyses, everyone can ride the more tempered hype wave toward widespread adoption without an unfortunate bubble collapse.
This story is published on Generative AI. Connect with us on LinkedIn to get the latest AI stories and insights right in your feed. Let’s shape the future of AI together!
