AI hype is overblown
There are many parallels between today’s AI hype and the lead-up to the dot-com bubble and bust.
Venture Capital firms poured $10.8 billion into AI and machine learning companies in 2017, according to PitchBook. A joint analysis from Pricewaterhouse Coopers and CB Insights reported $21.1 billion AI investment in Q1 2018 alone. And Google, Intel and Apple are racing to acquire more AI startups.
Piron argues that VC firms are irresponsibly driving overblown AI hype by focusing on large, flashy AI implementations. Sure, self-driving cars and image recognition services are impressive. But blockbuster AI systems, such as IBM Watson’s AI for cancer recognition, are notoriously unreliable.
“If you want to get an idea of the future of artificial intelligence, consider the internet mania of the late 1990s.”
Instead, Piron writes, we should focus on “less glamorous but also more reliable” applications of AI, such as the way Spotify uses AI to curate new playlists, or how call centers are using AI-powered natural language processing to improve customer service.
“Unlike many companies that fell into oblivion when the dot-com bubble popped,” he concludes, “the companies that have stuck around all took a long-term attitude. Those in the AI business need to remember this.”
Decades to go before AI matches humans
Google made waves when AlphaGo achieved a 4-1 victory over Go grandmaster Lee Sodol in 2016.
Speaking at the “Built to Change Summit” this June, a number of AI researchers poured cold water on AI hype.
Victor Muntes-Mulero, Vice President of Strategic Research at CA Technologies, had this to say:
“On true general intelligence there has been a lot of work, but the whole scientific community agrees that it will take a very long term before we reach it. It will take decades.”
According to The Economic Times, Muntes-Mulero believes we have “a long path” before AI lives up to popular hype. He points to a volume of ethical concerns as evidence, including sexism in Google Translate and the ambiguity of social rules.
Maria Valez-Rojas, a research scientist with CA Tech’s Deep Dive Research initiative, adds words of caution about “collaborative robotics“.
For example, she says, “When you tell a robot to ‘clean up the mess’, it might proceed beyond the coffee spilled on the floor and wipe out the massive calculations that you have done the whole week on the blackboard. Because that may mean ‘mess’ for it.”
Move past the mindless AI hype
If AI hype is dead, what’s left?
“Look for business use cases where AI is already a proven solution”, Jeff suggests. “And ensure that you have the data ecosystem available for AI to do its work.”
Instead, focus on a smaller, specific use case and manage your data wisely to realize AI’s potential for time and cost savings.
“Building a business case for AI isn’t so different from building one for any other business problem,” Jeff explains.
Before you sign any contracts, he says, make sure you’ve identified a need and desired outcome. Then, undertake a feasibility assessment. And finally, ensure that you’ve defined what “ROI” will mean.
In this article on Forbes, Jeff goes into more detail and outlines a handful of use cases for AI in Voice of Customer/Customer Experience Management:
- Creating better user surveys
- Reading and handling online reviews
- Pharmaceutical messaging and communications
Have some thoughts on the state of AI hype? Tell us in the comments below, or tweet us @Lexalytics!
In the meantime, stay tuned to our blog for more AI news and insights, interest pieces and thought leadership.