Artificial intelligence may be making people buy things
bbc.co.uk
bbc.co.uk
What's worth talking about in 2020 though is:
- What is the differential power that companies with advanced AI hold over other smaller players like mom and pop stores or smaller retailers? Is AI sophistication the new "income inequality" since smaller companies that can't wield that power get out-competed by larger players, leading to extreme concentration in the hands of few as time goes on?
- Compared to 2000-2010, what is the net increase in 2010-2020 for consumerism for someone making the same salary given online + AI influencing buying behavior? How does that fare for 2020-2030 and is that a sustainable path for the planet?
- How is AI influencing our buying habits and consumerism in general? Are we buying new type of things we'd never have considered before? Does it change who we are as people?
- Why are we not using AI to weed out bad reviews just like we do for email spam? Is the lack of AI in the right places equally harmful in making us buy more? Does Amazon really not have the resources to score reviews given how obviously fake top reviews tend to be for almost every non-brand item you look up on amazon?
This year it's AI. 10 years ago it was user analytics (well, it still is), and before computers it was having marketing department.
Large retailers with money to spend have always had a massive advantage over small retailers, both online and offline. AI is just the next incremental step.
That's not to say I endorse AI marketing, or larger companies outcompeting small ones. I just don't think AI is anything new in that regard, and after AI there will be some new technology (pheromones, I bet) that large retailers will be using to extract wealth from us.
Our experience as well. We built custom, turn-key, machine learning products for large enterprise for seven years now. One of our biggest frustrations is that our clients are exclusively huge companies because these projects are costly. I've had friends running companies ask me if we could do something for them, and I was embarrassed not to be able to help them build product because they wouldn't be able to afford our services.
So what we're doing to address that is to make our machine learning paltform[0] to reduce the time and effort it takes to build machine learning products. We feel that we can pull it off and we aim for 10x reduction in "time to value". Given that our business model was that of an agency with the mythical "engineer.day" unit, dividing that by 10 could make it affordable.
Second, amortizing work that was already done into services people could use. For example, we have forecasting functionality where people could forecast time series (say for sales, capacity, etc.) with seasonality, etc.
Third, teams such as ours could benefit from not having to make the mistakes we've made over these years and start at that level instead. These lessons cost a lot to learn in term of time and money. All the lost productivity to find out which dataset was used to produce which model, and what parameters produced which metrics. All is embedded in there.
There are many other reasons and objectives to do that.
- [0]: https://iko.ai
I don’t understand how this can be labeled news.
At least at the moment, I don't think this is a huge issue. I run a "mom and pop" business with my wife, and taking advantage of state of the art marketing technology through eBay, Amazon, Facebook etc. is dead simple and at least semi-affordable.
As the technology becomes more advanced, I don't see any reason why these massive companies would conspire to deny small businesses access to it. Sure, companies with in-house AI developers will have an advantage, but if it reaches the stage where this is make-or-break, small businesses failing will be the least of society's problems.
This is because you no longer own your relationship to your customer.
Sharecropping works but over time, you lose any leverage in the situation. You need them but they don't need you.
The end game doesn't involve anything good for you.
I work with many online stores that are switching from the sharecropper model after seeing the end game. They now have their own shops.
This, too, can be tainted by the big tech companies controlling access to customers but it's quite a bit safer. Repeat customers don't need to come again through the big channels and are often the real lifeblood of a retail business.
Regardless, owning your own shop doesn't magically solve all of your problems. You still need to bring customers in, be that through PR (which takes time), or advertising (which costs money).
What I'm really after here is to say that all those tools that the platforms give you come at a very high cost.
My point was that platforms are indeed hurting small businesses. We are now forced into a kind of surfdom. That's not a great future.
Your point about the platforms having too much power is right, but I don't think it's reached the level of full-on serfdom - at least not in e-commerce.
Sure, eBay/Amazon take a 10-15% cut of the sales, and this is arguably excessive, but they make it possible for product designers to sell their products direct to consumers on a large scale. Before they existed, the only way to reach a broad audience was through a hierarchy of retailers/wholesalers that might end up taking 70% of the final sale price.
It's not perfect, but it's definitely better than it was.
Usury, debt bondage, gig economy, aggregators. I think about this all the time.
The last stage capitalism flavored critiques hold no weight for me. I don't even understand what they're saying.
I want to hear about restoring and nurturing entrepreneurship, financing for cooperatives, fair profit sharing arrangements, community banking.
How did the Progressives and Good Government reformers talk about these issues leading up to The New Deal? I've been reading about those movements, but haven't gotten very far.
David Graeber and Clay Shirky make sense to me.
Recommendations, please.
I use an ad blocker online. I never see ads. I rarely shop on Amazon because of the counterfeit risk and overall mediocre satisfaction. When I do it's for something specific that I've researched in advance.
I don't watch TV and I have Youtube Premium so I don't see ads there.
I do most of my shopping at local retailers and buy the brands I've bought for years.
In that world even people who never see an advert will have only a hollow choice of products from the 'market'.
I share the sentiment of the OP; i am shielded from ads and even if they come through some way or another it’d influence me personally very little: i rarely buy stuff and when I do I research extensively. Some brands that are overly advertised are also a red flag to me.
But, I am aware I am in minority here. And those like us have a negligent impact at best. I think advertising does heavily influence the masses.
It's worse if it's just to brag about how advertising doesn't affect them, especially if the reason is that they have the technical chops to simply dodge the ads while everyone without those skills is deluged in AI targeted, psychologically optimized advertising they are ill equipped to avoid the effects of.
It's clear we are progressing technologically. Does anyone know of any science that describes how our emotions and the biology that generates them are progressing, particularly in our more and more technological world?
The article title seems a bit less obvious while also better describing what the article is about > How artificial intelligence may be making you buy things
Now we have word salad out of the way, I can say that it is absolutely possible to game the way people see the world to make them buy things they don't need or want and that GPT-3 has shown us it's possible to do that with a neural network which is basically pitting it against your average human's ability to resist it. This could be a type of addiction and corporations peddle that shit all day long as we well know.
In Germany, a Berlin start-up called SO1 is doing similar things with its AI system for retailers. It claims that nine times more people buy AI-suggested goods than those offered by traditional promotions, even when the discounts are 30% less.
To me it suggests exactly the opposite of the title. No AI secret sauce, just smoke and mirrors.
1. The title is meaningless. Replace the HN title's 'may' with 'may not' and it is exactly the same thing.
2. Especially if you consider the original title is How artificial intelligence may be making you buy things. There is nothing concrete in there.
3. It you read that quote again it is a very carefully worded and cherrypicked piece of data, something that is very common in AI hype articles. Without the actual numbers it is also meaningless.
Again, it is only my interpretation that the article is disingenuous, feel free to disagree.
No, the statement can be true and does not support the title. Say we have X AI offered goods and Y traditional promotions where |X| >> |Y|. Then it is not surprising more people buy AI offered goods. The statement remains true, but not what it wants to imply. Also we do not know other factors about the goods, say Xs are popular while Ys are not, they are not controlled for. Literally apples to oranges.
Let's blame the beneficiary instead, it's their purse and final decision. If we're worried about the environment, we should totally tax or outlaw bad practices, but do this at the appropriate level, not in software.
How does this compare to any loyalty scheme that people opt into? I'd think this applies also to people who pick up a paper stamp card.
Also the numbers are meaningless without a baseline.
I am suggesting that the quoted figures are what you would expect without AI being involved, and perhaps they are even not that good.
I think it's more a fundamental issue related to crowdfunding itself - you buy a product that you're interested in at the time, but don't receive anything until a year later. By then, there's a decent chance you've moved on and no longer need it.