This post may not age well if in a few years AI has become super intelligent, taken over the world and squashed humanity … but for now – yikes the over hype is real.
Seems like every company is promoting their AI this and AI that, even when it’s not really AI but just a computer program doing stuff it has done for years, or even when the “AI” is unnecessary or just plain bad.
I have been reading Billion Dollar Sellers which is an Amazon focused marketing newsletter and it has been breathlessly promoting AI tools (ie making money from paid ads / affiliate links) and promoting fear that unless a seller changes their product listings to be AI focused instead of keyword focused it is game over with no more sales.
For example, be more conversational in your listing to say your kids jeans are good for 5 year old boys who like to play in the mud because AI will use that to recommend your jeans to someone who asks AI for jeans for their five year boy who likes to play in the mud.
Okey Dokey.
First, that’s ultimately just using keywords.
Second, no one searches for products like that.
Third, even if someone searched for a product like that the number of searches are so ridiculously low (as in one) it would be a waste of time to spend limited and valuable product text on that search.
Fourth, even then this is a beginning search inquiry and not a buying inquiry. It is not going to generate sales – that is cash in pocket – compared to a more specific search such as where can I buy the size 10 boys Widget blue jeans.
Can sales be made from creating generalized buying guides? Of course. But what more typically happens is a buying guide recommends a product and then someone does a different search on that product to find the best place / price to get it – and that is where you want to appear.
Amazon doesn’t allow sellers to promote they are the best or have the lowest price so you’re not going to be able to target AI like this Amazon anyway. Maybe on your own website which then links to Amazon, although many sellers are better off selling directly from their own site instead of telling a customer to buy elsewhere.
A few experiences with AI
eBay AI
eBay’s AI will auto create product descriptions for sellers. Wowsa this is terrible. Hard to even call it AI unless you mean Almost Idiotic. It’s bad. Really bad.
Amazon AI
Amazon’s AI will now create brand names for sellers. Right now mediocre at best, assuming Amazon isn’t returning a response that the subject matter of your brand is blocked and no names will be created. Someone is really paranoid because huge swaths of product subjects are blocked.
Amazon will also create product information which is also bad although a bit better than eBay.
For buyers Amazon’s Rufus AI will create a number of questions buyers should look for before buying a product – topics which (1) Amazon does not even have a seller option to address in their product listings (oops), and (2) which are irrelevant to the product.
Even worse is when Rufus is using bad information from product reviews because many consumers are simply idiots or malicious trolls and their idiotic and trolling reviews then become gospel for Rufus about a product. Amazon might as well be telling people to shop elsewhere.
Facebook AI
Facebook will summarize comments about a post so you don’t have to read them. Amazon will do the same for product reviews.
I am find the summaries are so aggregated and generic as to be useless. Sometimes even misleading.
But this is probably good for stupid people who don’t read anyway.
I’m guessing on Amazon there are already Chinese sweatshops manipulating reviews
New Discoveries by AI
Finally, can AI actually do something new? Create solutions to a problem? Cure cancer?
AI does not seem anywhere close to that.
Summarizing previously created information to generate a product description or reviews may not even be considered AI.
Generating new graphics, based on graphics the AI has already been trained on, is just creating more of what already exists.
Ask AI to tell you if a company stock will go up or down next week and you’ll get a non-response and maybe a summary of past positive news and past negative news; maybe even a tip it will go up or down repeating what someone else is saying without making its own projection. In other words, again, just a summary of information from others. There is not an independent analytical analysis, that is actual Artificial Intelligence.
Although Grok by X seems to be getting closer to using existing information to make a future projection.
I gave X Grok and Microsoft Copilot some questions, such as:
will tesla stock go up or down next week?
who is more likely to win the next game between the denver nuggets and oklahoma city thunder
and Grok was miles better than copilot in (1) providing the information it was using to make analysis, (2) analyzing relevant information, and (3) actually trying to make it’s own projection versus repeating what was already listed on another website.
Microsoft Copilot seemed more interested in providing advertising links it could make money from than in providing an actual answer.
I had already removed Copilot from my computers and disabled it in Word and Excel. It can see some potential use in Excel, but otherwise its garbage and these tests didn’t do anything to change my mind that I don’t want it anywhere near my information.
Overall, that is why AI is way over hyped in its current form. Lots of hype, lots of electricity being consumed, little actual tangible benefits.
To be fair, I could have asked ChatGPT to write this post for me and it would have generated something in a few seconds. But it wouldn’t have my personal experience, examples, and specificity. It would just be general stuff summarized from what it “read” elsewhere.
Just to be sure, I asked ChatGPT if AI was overhyped and it responded:
“AI is both overhyped and underhyped, depending on the context.
Where AI is overhyped:
General intelligence: People often overestimate how close we are to AGI (Artificial General Intelligence). Despite rapid progress, current AI lacks true reasoning, understanding, and autonomy.
Job replacement: While AI is changing the job market, the idea that it will replace all jobs soon is exaggerated.
Magic bullet solutions: Some companies promise AI can fix everything—education, medicine, creativity—without considering real-world complexity or limitations like bias, hallucination, and interpretability.
Where AI is underhyped:
Productivity tools: AI is already revolutionizing software development, design, and writing, with tools like GitHub Copilot, ChatGPT, and Midjourney enabling faster, more creative workflows.
Scientific discovery: From protein folding (AlphaFold) to drug discovery and climate modeling, AI is genuinely accelerating breakthroughs.
Accessibility: AI has improved lives—think real-time translation, text-to-speech, and tools for the visually impaired.
So, AI is not a miracle, but it’s not just hype either. It’s a transformative tool with real limitations and massive potential.
Would you like a breakdown specific to a field like education, business, or healthcare?”
Frankly, after it said “AI lacks true reasoning, understanding, and autonomy” it should have just stopped and said it’s not really intelligence but a computerized data collector and way to quickly summarize information. Think calculator versus doing calculations by hand. Useful in some contexts, yes, but the over hype is real.
Link to this page