AI Hallucination Mushrooms
This week I came across three articles. All of them are about AI. They nicely illustrate the nonsense of using the word AI and using it as a substitute for true innovation and thinking.
The first is from the reputable The Wall Street Journal titled “Stop Worrying About AI’s Return on Investment.”
It starts with the line - “Tech leaders at WSJ’s Technology Council Summit said it’s nearly impossible to measure the impact of AI on business productivity. And when we try, we’re measuring it wrong.”
Truly astonishing admission. These tech leaders either introduce or allow new technology without any thought given as to why.
Here are a few gems from that conversation: “Return on investment has evaded chief information officers since AI started,” “CIOs are recognizing that traditional ways of recognizing gains from the technology aren’t cutting it,” and “racking up a few minutes of efficiency here and there don’t add up to a meaningful way of measuring ROI,” or “anything you actually care about you can’t measure.”
Obviously the good people from McKinsey had to chime in with wisdom like this: “One cannot expect significant productivity gains at the pilot level or even at the company unit level. Significant productivity improvements require achieving scale.”
And then somebody else sprinkled in the words of Oracle - “The other way of measuring AI’s return on investment is to use a top-down method, asking technology leaders to set business goals for the technology from the start, and then determining if AI—or some other technology—can help reach them.”
These are the words from technology leaders in their respective organizations. This not only puts into question their ability to be at the forefront of technology innovation, but at the same time it should be a gigantic red flag for all the vendors which are selling AI to them.
When people who should lead innovation in their company don't even know how to introduce technology in a meaningful way where you can measure the impact, you can sell them a shiny object once but then they lose any interest and start chasing something even more obscure.
You might see the MIT study “The GenAI Divide STATE OF AI IN BUSINESS 2025” where the authors state that “Despite $30–40 billion in enterprise investment into GenAI, this report uncovers a surprising result in that 95% of organizations are getting zero return.”
Any CEO and AI investor should ponder that.
The next gem was reported by Verge. The article title is 'You can soon attend Zoom meetings as your AI avatar' where the author shares the great news. You'll be able to create your look alike and attend meetings. Why would you do that? I have no idea. The next step will be to attach it to some AI which will be trained on your way of speaking. That on its own should give one pause. If AI can be so easily trained on the way you talk and write, how original are your thoughts... #timetojazzupmygame
Of course nobody ever asks the basic question - Why have these meetings in the first place? Only people who want to get fired shared these thoughts with the rest.
But the last article and video brought joy. A true example of innovation, thinking and using technology for the better. Let me introduce you to Forager HX800 - a robot that picks, trims and packs mushrooms. You can watch the video here and here. It is fun to see. It was created by a company called 4AG and the only unfortunate thing about them is that their website domain ends with '.ai'
As the tech is amazing, you should read the commentary from their CEO about the business:
“At first glance, replacing 80% of a worker’s output feels like you should also shed 80% of the cost. In reality, the tedious, time-consuming tasks that remain can carry a disproportionate share of the expense. Without true end-to-end automation, the ROI doesn’t hold up.”
“We could have pushed harder on sales and marketing then, but we also knew farms would quickly realize that those remaining 20% of mushrooms were still eating up ~50% of their labor costs. That ROI gap would have made adoption unsustainable.”
“Today, we’re closing that gap. On some farms, our robots are already harvesting 100% of the crop.”
The stark contrast between the pretentious 'technology leaders' who gifted us with wisdom like “anything you actually care about you can’t measure,” is that here you have a startup still under the same banner of 'AI' working on improving productivity. They know exactly what the technology should deliver and that ANYTHING UNDER 100% is not good enough.
The recurrent pattern? Some people pick up mushrooms and some just smoke them. Easy to spot the difference.