But google won’t give you the recipe. It’ll give you a pretty piece of text that resembles a recipe. You’ll only find out it’s not a recipe when it fails to produce a cake.
But then, the sites it's training on are starting to do the same thing, so maybe it won’t matter. Just last night, I pulled up four sites with “gluten free almond cake” recipes. One specified less than 1/4 the flour it would have needed, and another didn’t have any butter in the ingredients list. I had to eyeball the median and tweak from experience to actually get a bakable cake out.
I'm not going to drive to the bookstore only to find out the recipe I need twice a year is not there because the local cousine doesn't even know the ingredients.
Oh, maybe I should drive to the major city and check several bigger bookstores?
Or order it online? But that would be from the third country because this one is weirdly blind to variety. That will be €20 for a book and some €10 postage.
You're sure I can just get my recipe from the store?
Completely unrelated to AI, I buy and always bought recipe books in second hand bookstores. It has this vintage vibe, and not everything has to be kale. Downside is that some ingredients aren't easily reachable, but it doesn't happen that often.
I kinda think everyone should have an old school "joy of cooking", where they have instructions on skinning rabbits and squirrels, just as a reference that things weren't always like this not even that long ago.
Online grocery shopping service I use has added recipes to their website. Not obvious slop at first reading but then you see stuff like add 600g of carrots and 100ml of water to make a quite watery soup according to the picture.
The only solution is to find recipe books that were printed in previous decades.
Which is ironic, given that Google's entire value proposition (to users) was extracting signal from noise...
... and now it's come full A/B-advertising-optimization to being useless at that, when the need is greater than ever.
Imho, Google's greatest failing was missing how its own incentives warped creation of new web content, and failing to account for that strategically -- it turned the web into something it can't itself usefully parse.
We have a distinct poverty when it comes to secondary considerations and long term ramifications - which used to be manageable when progress was slower. Now, we're on a very steep acceleration/progress curve, and any shortsighted mistakes cause extremely large ramifications. Which are then compounded by both more short sighted non-fixes and our rapid acceleration/progress curve layering in additional confusion, misunderstandings, omissions of critical information.
What kind of insanity are you all talking about? Plenty of cook book authors out there with good recipes, tested, big-personal brand, etc. no need to go to prior decades for good books.
You need to compensate for not adding it though. The recipe without was literally the same ingredient list as the site it copied, just missing a line.
And almond flour does its thing by carmelizing in combination with butter and sugar, turning your whole cake into a sort of giant macaron. You can’t pull it off without any one of those things.
> But ultimately that strategy is good for the consumer right?
No. It may be good for the consumer right now, but not ultimately (and on top of that, I would argue that reducing everyone to a consumer is already the wrong framing -- you need to ask what's good for the citizens). Having ten competing supermarkets with various interweaving supply lines is ultimately much better than having one giant supermarket, because that one monopolist is able to squeeze both consumers and suppliers to the detriment of both.
Do you think Walmart’s prices are higher than you’d pay if there were just ten competing stores with the same product mix, each of which did not have the leverage to squeeze suppliers?
It doesn't matter what Walmart's prices would be -- because there'd be nine alternatives offering their products at different price points. Also, you seem to be implying that suppliers are inhuman and deserve to be squeezed (but Walmart somehow deserves all the wealth it extracts as a middle-man, go figure), which is exactly why I said that focusing on consumers only is a myopic view of the world.
Without some way to generate revenue, people aren't going to publish recipes (for Google to scrape into their AI.) Maybe we could live without more recipes being fed into the machine, but there are many other types of content that will suffer the same fate.
It would be nice to find something better than an ad-revenue driven web, but I'm not sure this is it. We'll find out I guess...
Those who won’t were doing it for the money. Those who continue are those who do it for passion, or those whose recipe is just a way to attract people to their business (e.g. kitchenware company). I don’t think it is necessarily bad, the quantity will decrease but the quality may even improve
I built a passion reference site. A large part of that passion came from knowing and talking to the people I was helping. One person emailing or saying thanks would later help power me through to create more useful articles. Enriching openai/claude/ms/google and no thanks from an individual, has disincentivized me from writing more.
Same here. People knew the website and it was immensely flattering to meet users in the wild. It motivated me to really sweat the small stuff, because people noticed. Now I'm just feeding the slop machine, and it feels pointless.
We just won't get countless recipe websites where you have to scroll, scroll, scroll through slop about someone's day to read a scraped recipe that every other website has.
Does the current Google search results indicate that they will be any different?
It’s also not disruption if your product relies on the output of the industry it kills. What will AI train on when it destroys the economics of sharing information with others?
From the HN guidelines: "Please respond to the strongest plausible interpretation of what someone says, not a weaker one that's easier to criticize. Assume good faith."
In this instance, the internet is more than top-ranking recipe websites.
> But ultimately that strategy is good for the consumer right?
No because it's killing competition and becoming an even more obvious monopoly. Then at any occasion they have to choose between consumers and profits, they'll do what shareholders want and increase profits.
It's not uncommon for free things to be higher quality than cheap things, especially when we're not talking about physical goods. Think hobbyist vs hack. Selective pro bono vs quantity over quality. The former describes old internet while the latter describes much ad-supported internet. I'm not saying cheap is better than expensive, and I'm not saying everything works this way, but I do think many things do, especially for pure information that doesn't have a major capital cost associated.
Not really good for consumers in the long term.
Creates a monopsony (monopoly on the buyer side instead of seller side) in the supply chain, with all the same negative externalities to competition.
Same reason its not good to have a "company town" where 1 company is the major employer to 50% of the workers.
Not even the summary of the recipe, I think a lot of the "value" AI provides is decluttering the interface. All these perverse incentives conspire to make it undesirable for web publishers to simply show you a cake recipe. They need to meet SEO metrics for being higher on googles index, they need to monetize with ad and chum boxes. They need to distract and dazzle your senses. In the end, AI strips all these anti patterns out and shows you the meat. Or the "summary" of what should have been there all along. The "AI" isn't learning about what the words mean, it's stripping out cruft and filler that was deliberately added!
I'm not saying LLMs are worthless, but I'm saying if you had a magic browser add-on that simply stripped the BS out and showed you the relevant content, it would meet the use cases of the majority of people using AI (regarding search).
Said differently, Google are bringing you a solution to a problem they (largely) created.
except this isn't a solution. taking away sources just creates another problem down the road. there's no information to crawl from if no one is building websites to crawl from.
