The US spent a total of $130.6bn (£98bn) between 24 January 2022 and 31 August 2025
So, a fraction of what one single country of ~25M people is giving the USA: https://www.abc.net.au/news/20...
I suppose it's challenged by defining a dishwasher cycle (maybe an average power draw over 100W for 30 minutes on that circuit) and then figuring out when i last bought them and the size of that pack. Then it's one thing it work on Amazon, but it should really scrape everywhere that can deliver.
It's totally something I could build since i'm tech savvy, but it's not worth me investing a day of my time to save on dishwasher tabs. But being able to describe an agent in plaintext and have it figure out all the pieces - that could be really useful.
That would be the "fun" way for a tech-savvy person to solve the problem. But the "easier" way would be for the system to know how many dishwasher tablets (or whatever item) you have in stock. When stock falls to a certain threshold, trigger the algorithm to go scrape some websites and find the best deal to buy more. This could be solved onsite by scanning barcodes when you are putting groceries away, or by scanning the receipts.... but of course our overlords at Amazon already HAVE this information. A physical store has it too if you used a loyalty card when buying the items. Visa may already have the information too, depending on the exact nature of your credit card agreement. If you pay by phone/smartwatch, Android or Apple might have the information too. If you use instacart, they also know what you bought. All of the above businesses may also enter agreements to sell/share this info with any other entity that is profitable.
So, it's probably easier for any of those big rich companies to build this tool that knows all about you, than it is for you (even as a tech-savvy person) to build it yourself. Those multiple big rich companies already have all the information about you and your purchasing habits, we're just waiting for them to decide to let it work for us, or, more likely, to dress up even more data-mining or advertising in a veneer of "working for us."
You probably know all this, I'm just putting it out there as a reminder for others
Well yes, culture is the problem. But not necessarily just a pro-gun culture. Canada and Switzerland have that, too.
It's a "I'll solve every conflict with violence"-culture and "Look at me and I'll shoot you"-culture that is the actual problem. Look at the US current politics. And - as you mentioned - a "Well, too bad we can't do anything here. Thoughts and Prayers"-culture
Pro-gun-culture is merely a big enabler.
Do Canada/Switzerland have stand-your-ground laws or Castle Doctrine? Genuinely curious.
until AI can have sex for me.
I hope that any business that tries this will get totally fucked. Does that count?
And it's often the fault of the local businesses. If I want to buy "unsalted mixed nuts", my local store doesn't carry them any more. (It used to, but it stopped.) So now I buy them on Amazon.
Did you ask at the store? I've had plenty of success doing this. Or even if that store can't help, they will often know of another local place that *does* carry the item. Sometimes they'll even phone the other store to confirm for me.
If i can control it, then i think it'd be great.
Like I'll sometimes buy dishwasher tablets on Amazon - the price has to be less per tablet than costco, and we need to be most of the way through the current box. An AI could in theory be on top of that - and I'd be totally happy to offload that labor. In theory that was kind of the dream of "subscribe and save" but it's so unpredictable and I miss so many of the "your subscription went up by $19 but you're still getting the best price!" emails that I pretty much gave up on it.
Even by the current diluted definitions, that doesn't sound like "AI." It's a simple IF THEN loop, and that sort of thing could have been implemented for online shopping decades ago. The fact that it hasn't probably means it's either a dangerous idea, or it's too friendly to the customer, so no business wants any part of it. The notion that any current "AI" would act in the customer's favour is similarly absurd. Visa can fuck off.
Literally an entire website dedicated to geek gifts existed for 20 freaking years. And they think this is only now becoming a thing?!
And they couldn't stay in business. So, to answer the question in the headline, no.
Betteridge's Law of Headlines holds firm.
Man, I seriously miss (the old) ThinkGeek.
+1
So you would be prepared to fly to Tasmania and take the donkey caravan to the remote village where their office is located to deal with this matter?
No?
What about if the shoe is on the other foot... what should the chap who lives in that remote village do if he needs assistance from a physical office in New York City?
At least it would be an option.
This is every company that uses any kind of automated system for this sort of problem. Pray you never set off any of Google's automated systems because the odds of getting that fixed are just as bad.
Companies should take this more seriously as there's nothing stopping a bad actor from intentionally inflicting these sort of problems to damage the company's reputation. If someone could make enough money shorting stocks to take advantage of bad publicity around the problem there's sufficient incentive to engage in it.
Agreed. If I were dictator of the world, I would mandate that any online business with more than some number of users would be required to have physical office. So if a user's account got cancelled, as a last resort that user could physically go to the office with their physical ID and get it fixed.
Yes, this would mean that some online businesses were no longer profitable. I still think it's worth it. Especially compared to the systems we have at present where the user uploads selfies and pictures of their IDs to god-knows-where online.
The problem with this is that google is often the alternative and can be just as bad.
There's a documented case out there where during COVID, when doctor's offices were mostly shut down, where the parents ended up taking a picture of the baby's rash (yes, in that area), and emailing it to the doctor's office.
Somehow Google's automated child porn detection systems flagged this, it was reported to the police who opened an investigation, then closed the file when it was realized that: 1. It was sent to a pediatrician office, 2. Yes, baby had a rash, 3. It was COVID, so office working remotely. This worked long enough for the family to get the appropriate cream for the baby, but google also closed down the man's accounts and marked him as a nasty banned pedophile.
I never did hear if he got that resolved.
Didn't Apple try something similar with their cloud photos? IIRC they were going to scan all of them for CP but the outcry from parents made them walk back the decision.
It didn't tell me what to select. It did a side-by-side comparison of the various attributes of each plan, comparing things like deductibles, coinsurance, max out of pocket, etc. While I don't doubt that Gemini and its kin will offer product placement, it's a whole other thing to distort such a comparison.
I'm not saying it told you what to select in your case, I'm just saying it would be a fairly small step if Gemini's owners ever decide to do so. Of course, it wouldn't be so blunt as to say "buy this one." The summary would just use biased words when describing the prices, or exaggerate the uncertainty of the potential costs of a given plan
At first glance one might think "numbers don't lie" but I'm sure we have both seen enough advertising to know that numbers can be misused to promote one view over another, or to try and convince customers to accept a huge price hike. When the advertisers get their claws into AI, they will do this, but with misleading figures custom-generated to mislead each specific person.
To be clear, I did say I was being pessimistic. I also didn't say that in your specific case the summary was gamed to manipulate you into doing something. But since writing the original post, and this one, It is becoming clear just how straightforward it would be to subtly manipulate an AI summary to make sure that the customer chooses what the advertiser wants them to choose.
Note that Gemini is owned by Google, who make huge amounts of money from advertising. Especially personalised advertising, customised based on the huge amount of information they have about you because they have such huge penetration into web browsers, smartphone operating systems and email. I cannot imagine that they will just leave the AI advertising money on the table in the future.
I used Gemini in Chrome just yesterday.
My son was signing up for his company's health insurance plan. He had three choices, and the options weren't easy to decipher, and the text was long. AI was able to show me a nice three-column comparison of what was different in the three plans, just by reading the text on the page. I spot-checked its findings, they were right on.
Despite the backlash from many, there are many ways in which AI helps people in very tangible ways.
I'm probably just being pessimistic, but it seems like a very small step to have the AI in cahoots with the insurance companies to make sure the summary tells the customer to select the plan that is most profitable for the insurer.
Someone needs to train the AI to click around the web like a user, so that the web companies can keep their ad revenue up once they've knocked us pesky humans off the web.
There used to be a plug in for that. At the time it was intended to obfuscate the user's actual browsing habits, and to make any data collection worthless. It would operate in the background and "click" on random links so that nobody could tell if ad views etc were real.
This industry is a problem the world over.
Fortunately there is a way to fight these businesses. Just don't buy their product. This will mean that you simultaneously save money and reduce their revenue. It's not exactly a "life hack" but it's true.
The more data I punch in this card, the lighter it becomes, and the lower the mailing cost. -- S. Kelly-Bootle, "The Devil's DP Dictionary"
