Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

I used AI-powered calorie counting apps, and they were even worse than expected

lifehacker.com

Post reply on HN
Print view View original

261–270 of 302 posts

Re: I used AI-powered calorie counting apps, and they were even worse than expected

#261

Post by malfist »

mminer237 wrote: post #256
malfist wrote: post #252

Earlier quoted context omitted.

How can your AI hope to be accurate. Visually, Can you tell how much butter or oil someone used in the recipe? Visually, Can you tell if it's 73/36 ground beef or 97/3? Can you tell if that taco meet is beef or turkey? Can you tell if that glass of milk is skim or actually just heavy cream? You can't tell if I cooked my scrambled eggs in the rendered fat from the sausage I cooked just before it, or if I used a teaspo…

You can just assume an average amount of butter for such a recipe, 80/20 beef, beef, and 2% milk and be reasonably accurate, especially for people who otherwise wouldn't have any idea of their calorie intakes.

What is "an average amount of butter"?

And why can you assume any of the other things?

Re: I used AI-powered calorie counting apps, and they were even worse than expected

#262

Post by gadders »

wredcoll wrote: post #255
gadders wrote: post #200

Earlier quoted context omitted.

In your early 20s your metabolism is that much higher as well that it can cope.

Oh so that's why college students famously gain weight when they go to college, they're too old to have a fast metabolism!

How much more would they weigh if they underwent the same changes at 40? It's faster metabolism, but it's not magic.

Re: I used AI-powered calorie counting apps, and they were even worse than expected

#263

Post by boydYVR »

foxyv wrote: post #244

Calorie counting is so unreliable it hardly even matters. The calorie measurements are way off. There are so many sources of macronutrients and they are all digested differently and lead to so many different amounts of "Fuel" for the body. Just the different types of carbohydrates are bad enough. An actually accurate food label would look like the spec sheet for a 5th generation fighter jet. This doesn't even take in…

Yes, all bodies process food differently, and all measurement tools are inaccurate. One of the bigger benefits of tracking is the ability to see trends. If you're chronically low or high on a nutrient, a good tool will make this apparent to the user.

Re: I used AI-powered calorie counting apps, and they were even worse than expected

#264

Post by sunnyam »

cowanon2222 wrote: post #31

I've had pretty good results using the AI features in Macrofactor. It's certainly not perfect, but it does a pretty good job with mixed text and photos and allows you to easily fine-tune the results. Macrofactor is also the only app I've seen that actually estimates your underlying metabolic rate and adjusts accordingly. It predates the recent AI surge, and seems to have a team that's studied nutrition science behind…

I think the key difference is that they perform a search of the foods in Macrofactor's database which means that you're more likely to get a good estimate.

From someone who weighed and scanned a lot of foods, it has really improved the workflow

Re: I used AI-powered calorie counting apps, and they were even worse than expected

#265

Post by Kirby64 »

gadders wrote: post #262
wredcoll wrote: post #255

Earlier quoted context omitted.

Oh so that's why college students famously gain weight when they go to college, they're too old to have a fast metabolism!

How much more would they weigh if they underwent the same changes at 40? It's faster metabolism, but it's not magic.

20s to 40s metabolism isn't that different. 4-5 pints is nearly 1000 calories, and basically nobody has a 1000 cal decline in base expenditure from 20 to 40.

A bigger change, generally, from someone in college and someone in their 40s is their activity levels. Even just considering the amount of walking most folks do on a college campus is a huge difference, compared to someone that gets in their car and drives to and from work.

Re: I used AI-powered calorie counting apps, and they were even worse than expected

#266

Post by hammyhavoc »

malfist wrote: post #261
mminer237 wrote: post #256

Earlier quoted context omitted.

You can just assume an average amount of butter for such a recipe, 80/20 beef, beef, and 2% milk and be reasonably accurate, especially for people who otherwise wouldn't have any idea of their calorie intakes.

What is "an average amount of butter"? And why can you assume any of the other things?

Probably because most people likely to use this app aren't going to be using it for serious self-monitoring like someone with diabetes would. The reality is that most people will use an app like this for a while, fail at their attempt to lose weight or gain muscle, and lose interest. "Good enough" is the mantra of much of what is in the health/nutrition/fitness niche and it goes beyond anything AI-adjacent.

Re: I used AI-powered calorie counting apps, and they were even worse than expected

#267

Post by hammyhavoc »

ericmcer wrote: post #224
wnorris510 wrote: post #183

Founder of https://snapcalorie.com and ex-Google AI researcher here. The article actually gets our pricing wrong, it's completely free to use SnapCalorie, the $79.99 price is a donation to support our research and help us continue providing the best tools to everyone for free. We've published in top peer reviewed academic conferences and our algorithm (can't say the same for other apps) IS actually more accurate than…

Thanks for mentioning voice, I prefer that to photos. Photos seems like the most complex technically. AI seems to be going heavy into text/chat and images but voice is really the perfect UI to me. I would be totally fine if I could trigger a "call" from the app and talk to an AI on the phone for 30s and it could log my meal, give me some advice, etc.

Do you talk to your voice assistant on your devices as-is? Do you do it in the company of others?

With the exception of a few legally blind folks I know, I don't know anyone who openly talks to a machine, nor anyone who does it in private.

Re: I used AI-powered calorie counting apps, and they were even worse than expected

#268

Post by n4r9 »

gadders wrote: post #200
reedf1 wrote: post #128

Earlier quoted context omitted.

The idea that a night out with 4-5 pints might clear my daily total requirement for calories was not something I had really considered until my late twenties.

In your early 20s your metabolism is that much higher as well that it can cope.

That doesn't match my experience, or most of the stuff I've read. The number of calories I need to hit to lose weight has been pretty constant from my early twenties to late thirties.

That said, I didn't have to be as careful in my twenties because I did a lot more exercise. And that's because I had more free time and opportunity for sports, fewer energy demands, less money for food, and more incentive to walk or cycle places. So I agree it's probably easier for university students to be slim, but I suspect metabolism is not a primary reason.

Re: I used AI-powered calorie counting apps, and they were even worse than expected

#269

Post by mapt »

jonplackett wrote: post #211
mapt wrote: post #203

Earlier quoted context omitted.

I wanted to examine this numerically. * Great Value Original Vanilla Lowfat Yogurt, 32 oz Serving size 3/4 cup (170g) 1.5g fat 26g carbs (21 of which are sugar) 5g protein 130 kcal * Great Value Greek Plain Nonfat Yogurt, 32 oz Tub Serving size 2/3 cup (170g) 7g carbs (7 of which are sugar) 17g protein 100 kcal * Great Value Light Vanilla Nonfat Yogurt, 32 oz Serving size 3/4 cup (170g) 15g carbs (12 of which are sug…

https://www.sainsburys.co.uk/gol-ui/product/fage-total-0-nat... This is the yoghurt I was talking about. It’s a high quality brand here in the UK 54kal per 100g and 10g protein 3g carbs (zero of which sugar) Ingredients: Pasteurised Skimmed Milk, Live Active Yoghurt Cultures (L. Bulgaricus, S. Thermophilus, L. Acidophilus, Bifidus, L. Casei)

My mistake. This is very close to the greek yogurt I mentioned, but the serving size is different - I was using 170g recommended serving size instead of the standard 100g comparator. Also - all 3g carbs are sugar (naturally occurring).

Re: I used AI-powered calorie counting apps, and they were even worse than expected

#270

Post by mgraczyk »

wnorris510 wrote: post #183

Founder of https://snapcalorie.com and ex-Google AI researcher here. The article actually gets our pricing wrong, it's completely free to use SnapCalorie, the $79.99 price is a donation to support our research and help us continue providing the best tools to everyone for free. We've published in top peer reviewed academic conferences and our algorithm (can't say the same for other apps) IS actually more accurate than…

What do you mean by this?

> it's completely free to use SnapCalorie, the $79.99 price is a donation to support our research

I just downloaded the app and I don't see any way to use the premium plan without paying 79.99 or 6.67/month. How do I choose not to donate?

Or do you mean it's a required donation? And if that's true, in what sense is the article wrong, given that it says exactly that?

Post reply on HN