AI Hype: Bad Data Is Bad Data.
It's meaningless without a Y axis!

AI Hype: Bad Data Is Bad Data.

New data highlighted in Casey Newton's Platformer newsletter codifies what most of us have already assumed: AI chatbot usage is starting to reshape search. And when search changes, so does the Internet as we know it. Unfortunately, the data lacks a fundamental denominator, and as such, only serves to feed the signal-free hype cycle we're currently in.

The data comes from Adobe's Analytics platform customers, and it paints a fascinating if incomplete portrait of how consumers conduct their online research. Yes, traffic from AI chatbots has risen more than 1000% since last summer, but....on what base? Take a look at this chart:

Note the Y (vertical) axis - it's in percentages, not in real numbers. If there was one visit in July of 2024, and 17 more folks visited last month, well, there's your 1700% increase! If the base were a million users, that'd imply 18 million last month - significant, but still not that big. Without knowing the base, the figures are meaningless.

These kind of "numbers go up" pronouncements are typical of an industry caught in a hype cycle - Adobe sells a lot of AI-related products after all. And it feeds a press eager to prove pre-conceived notions of AI's indisputable rise. But I find it telling that nothing in the data released gives us a baseline from which we can truly understand generative AI's impact on how search has traditionally driven traffic around the web. And that's a shame.

?

?

Harry Fozzard

Growing global organziations with full-service digital marketing.

3 天前

And yet, AI is definitely a thing that will reshape more than online search. What would be really interesting is to plot the increases in power/utility across areas that AI will reshape: biomedical research, mid-level management, robotics/manufacturing, etc.

回复
Pushan Banerjee

E-commerce | Digital | Strategy

1 周

All aboard the AI hype train! The problem with most data analysis is the fact that the analyst knows the reader doesn't know how to interpret data. That this is from Adobe means this is meant to draw C level people into investments that they have no idea about

要查看或添加评论,请登录

John Battelle的更多文章

  • Can Google Win In AI Search?

    Can Google Win In AI Search?

    Two years ago I wrote a series of posts exploring the business model and interface implications of generative AI-based…

    1 条评论
  • Why Haven't AI Agents Happened Yet?

    Why Haven't AI Agents Happened Yet?

    Three months ago I published my annual predictions, and while I rarely revisit them in the middle of the year, I do…

    7 条评论
  • When Did Tech Stop Being Magical?

    When Did Tech Stop Being Magical?

    I’ve been pondering something for a while now, but have held off “thinking out loud” about it because I was worried I…

    25 条评论
  • What Are You Reading, and How?

    What Are You Reading, and How?

    Nearly every conversation I've had over the past month has involved some variation of this question: What are you…

    3 条评论
  • Tech Has Replaced Finance As Too Big Too Fail

    Tech Has Replaced Finance As Too Big Too Fail

    I opened my annual predictions last week by noting that the technology industry had leapfrogged finance as the most…

    2 条评论
  • 2025: The Year of the Big Tech Flex

    2025: The Year of the Big Tech Flex

    This isn't going to be a normal year. 2025 will be strange, frenetic, and full of surprises, particularly for those of…

    7 条评论
  • Grading My 2024 Predictions

    Grading My 2024 Predictions

    2024 is in the books, so it’s time to grade my own homework. One year ago I posted my 2024 predictions, fresh off a…

    3 条评论
  • Bluesky Is Getting Big. Does That Mean Advertising Is Coming? (Yep).

    Bluesky Is Getting Big. Does That Mean Advertising Is Coming? (Yep).

    I’ve been in the business of making new kinds of media companies, media platforms, and media technologies since before…

    2 条评论
  • Generative AI Won't Work ... Unless We Change Our Approach

    Generative AI Won't Work ... Unless We Change Our Approach

    Listen up, tech oligarchs; lend an ear, simpering brohanions. We’re doing this generative AI thing all wrong, and if…

    2 条评论
  • Why BlueSky Is Taking Off

    Why BlueSky Is Taking Off

    Emily Liu at Bluesky has a timely post that I'd like to respond to. (Back in the day, when blogging was a thing, we did…

    5 条评论

社区洞察