How AI Was (Briefly) Hijacked by Hype - Why We Can't Have Good Things
Hypothesis on how an intelligent new technology was overtaken by hype from covid-era business school graduates, uneducated marketers, and influencers- along with how it could thrive again in a new era
AI hype grew week over week in 2024 and peaked; encounters in past week alone:
A Nigerian man in a tuxedo asking me for an AI job
An Australian surfer-turned-AI-guru
Recruiters pinging me to join companies founded by influencers
My background:
9 years of traditional ML and DS in Silicon Valley, recent AI consulting for major corporations, and the first AI strategy work between C-level legal and privacy teams at Fox Corporation (I left as my google doc collaboration style clashed with the Fox style: old men fighting over budget for software applications marketed as AI).
My hypothesis:
Current AI hype is fueled by two key groups lacking critical thinking (there is more, but these have been front and center to me):
Covid-era MBA graduates
When researching the least-thought-out of the startups I’d encountered, a common theme was recent MBA attendance (unless they were from Bangalore - no themes there other than being from the same location). I began to think their founders’ practical experience or humility was likely compromised by lockdowns - this is only based on anecdotal evidence bias but also due to global reports of where this cohort went.
Uneducated marketers
If this represents the common visual of AI literature, then you’ve likely been reading about AI through marketing by marketers with superficial understanding.
And Marketing (without a data team to measure) Includes Influencers
Obtaining millions of followers despite dubious qualifications, funding seemed driven by marketing teams without a qualified data science team (or data science teams who received an ‘online certification in 2020’ and were employed by others with same cert), it perpetuated the image-seeking mentality of sharing hype through emojis.
My blog post on this was here and examined correlation between # of emojis per overexcited AI-themed LinkedIn post by HuggingFace employees and Nvidia stock price.
Technology and where we go from here:
AI's advancement requires time for problem-solving and accuracy, rendering many hyped applications (e.g., AI drive-throughs, autonomous customer service) impractical while opening doors for innovative solutions.
Unfortunately AI may face a “scam” label from disillusioned investors and public - while I hope not, it would require a recalibration period, during an election cycle, so I’m not super optimistic anytime soon but would love to be surprised.