Hacker News Reader

Comment by DantesKite | Hacker News Reader

It's ironic. Software has affected almost aspect of our lives, except cooking.

You have all these recipes, but no real great way to determine the best version of something.

There are certain techniques (simple techniques) that'll make steak and coffee taste better, but they're hidden amidst a vast empire of information and useless knowledge.

There's another aspect that people rarely think about:

100 years from now, will humans still be cooking food? Or will that be regulated to machines?

If so, you'll need a way to keep track of all the recipes and what humans enjoy the most. You need a way to filter what's good from what's great.

And when you zoom out, it's a problem that's common to GoodReads, Netflix, and information in general.

How do determine what people will like? And how do you grade it?

View on HN

>100 years from now, will humans still be cooking food? Or will that be regulated to machines?

Of course we will. Cooking is not just enjoyable, but a unique expression of someone's personality.

I come from a food-oriented Italian family and your entire comment just blows my mind.

I'm also Italian (from Italy, not "my great grandpa was half Italian"), if some robot can make me parmigiana di melanzane as good as I can make it, I'll never cook it again.

You’ll spend the same the same amount of time cleaning it instead.

I'm French. I find cooking relaxing.

We enjoy the results of cooking: the food, the experience, the compliments we get.

We also enjoy improving our skill.

But we would be happy to have a mechanical sous-chef who will do all the work and let us focus on the fun creative parts.

And that's only if, there won't be a more enjoyable, activity to do.

And for most meals, we would even prefer not to bother at all.

Look up the Thermomix. It does a lot of those things for you automatically.

It also has a manufacturing fault that gives you third degree burns and the company tried to avoid recalling it, but there's almost certainly a good competitor product out there.

I should've put it another way.

I imagine a hundred years from now, an overwhelming amount of the time, machines will cook food for humans just because it's easier.

But I imagine people will still occasionally cook for themselves, the same way people still play chess against other humans.

Don't we already have that, in the form of the numerous food delivery services? You can log into your "machine" (ie. phone) and order whatever you want. It then magically appears at your front door. It doesn't pop out of a door in the front of the machine, but isn't the effect the same? The "machine" hasn't eliminated cooking today and I doubt that it will have in 100 years.

Personally I've never used an Internet based delivery service, as I prefer home cooking for a number of reasons. I suspect that ease is only a minor factor in food choice.

Food is also complex, in that there are so many combinations. Given 50 possible ingredients, there are a huge number of possibilities (plus interactions between ingredients) if you choose 6 of them to make a meal. How do you capture that in a machine, even neglecting that cooking technology continues to evolve?

You're probably right - Uber Eats et al have already shown that there's demand for this even at high prices.

Yet another simple human pleasure sacrificed at the altar of efficiency...

(To anyone in Australia or with an Aussie VPN, there's a wonderful Bluey episode about this here: https://iview.abc.net.au/video/CH1903Q034S00)

Bluey is great, and as a parent it is relatable in so many ways. But I don't think your link points at a particular episode. At least, not for me.

Name of episode?

My apologies. Bin Night. Classic Bandit episode that touches upon the importance of the mundane things we do.

It's bad enough Netflix algorithms try to shove things in my face whether I like them or not, I don't want the food equivalent doing it literally.

> You need a way to filter what's good from what's great.

Existing algorithms already struggle with being contextual. On youtube/netflix I often want short funny videos, but other times are better to sit and watch that four hour long history of the parthian empire but the algorithms have no context for this and give me the same suggestions at both times. Food would be even worse, different meals for breakfast lunch and dinner, a desire for variety and some deeply subjective preferences would make it all but impossible.

Another issue is that cooking and taste co-evolve together. You could get the worlds best chef to make a meal but it will never be as good as me making the same meal that I've refined over the years and that my taste buds have got accustomed too. For me some of those recipes split off from my Mothers and may have diverged from what everyone else considers normal.

I think this is just part of the human experience we need to keep part of the human experience and just let the robots do the cleaning up.

Has anybody seen a working meal recommendation algorithm? Looking for new dishes which are still appealing to my entire family is a constant struggle.

Hmm...

I think nuance and novelty are too significant in the cooking world to ever allow machines to take over 100% of cooking unless they offer incredible levels of flexibility. We tend to get bored of even the "best" flavors and the "best" is highly subjective. I can't speak for everyone, but feed me my favourite food, even my 5 favourites and I'll be sick of them all by day 10.

Cooking has been massively impacted by software. Even if you’re only talking about home cooking.

The shared information available is truly phenomenal. Sure there’s lots of low quality cooking information, but that’s the same in all domains. It’s many orders of magnitude superior to the pre-internet level of information available.

The ability to sift through all of the information out there, including an endless amount of user generated content, and have it translated from basically any language you want is incredibly useful. Being able to find a video tutorial for almost anything on demand is incredibly useful. Being able to search for shops that stock exotic ingredients is incredibly useful. FaceTiming mom to get her help with a technique is incredibly useful. Having a private GitHub repo for sharing recipes with my friends is incredibly useful.

I’d completely reject the idea that the information revolution hasn’t been hugely impactful on cooking, even if you still have to put in the effort of trying things out for yourself.

Maybe I was being a bit too hyperbolic.

I'm just frustrated at how much better cooking could be.

You're right. It's much better than it used to be.

I just dislike being stuck wandering around the Library of Babel.

I’d suggest that every field of knowledge has/is approaching a Library of Babel type situation. I think the way to navigate it pretty much always boils down to practice, experience, and all the other typical learning behaviours. If you want to follow that path with cooking, you have to really enjoy the process though. Because putting tremendous amounts of effort into home cooking is probably the least efficient way to nourish yourself.

Another problem that I think the field of culinary knowledge has, that a lot of other fields also have, is the prominence of the highly opinionated knowledge sharer. Especially if you’re interested in traditional or culturally significant dishes. Most people will want to try and convince you of the one true way to make something, when really they’ve just got the particular way that they make it. If you want to get past that barrier, you have to take a look at as many different perspectives as you can find, to try and seperate the core characteristics of a dish from peoples individual idiosyncrasies, and put aside the dogma altogether that there even is one truely correct way to make something.

It's not that hard to find great recipes, from respected chefs with good user reviews, that have high likelihood to turn out very tasty.

So how could you improve that? And where's the value?

Yeah, I’m not sure what you’re expecting. There is no substitute for trying the recipe. What do you expect the software to do?

I’m reminded of articles about coming up with the “best” chocolate chip cookie recipe. I found a good one [1] though it’s not the one I remember.

Each recipe is a point in an n-dimensional space, and that’s just considering the amounts of each ingredient.

https://www.seriouseats.com/the-food-lab-the-best-chocolate-...

I'm not sure to be honest. It's a hard problem to solve. I don't know if it'll ever be solved.

"100 years from now, will humans still be cooking food? Or will that be regulated to machines?"

The wealthy will. I've noticed my wealthy neighbors hire help for everything. Stuff that I don't mind, they hate.

I walk by their empty garages with nothing besides a nice car.

They don't seem to have any hobbies, or interests, besides their work.

They hire help to cook. The hire help to rearrange their furniture, or rearrange their closet. They hire nannies when a parent is not working. They let strangers clean their home.

As a poor guy I like cooking, driving, and even cleaning some days.

Eddie Murphy's bought a nice Porche years ago. In an interview he stated he loved washing, and polishing it. I think that's the difference between the wealthy, and the rest of us?

I enjoy cooking, and even driving, most days.

I have never understood my wealthy (Marin County, CA) neighbors.

Reply on news.ycombinator.com