Anyone who has participated in calorie-tracking for an extended period of time can tell you that the visual appearance of food can be highly misleading to the number of calories in it.
I'm all for making it easier for people to lose weight but this app may honestly have the reverse effect. If the app estimates calories too low (and therefore the individual eats more), many people will get frustrated with the lack of progress and give up. If the app estimates too high, the individual will lose weight, but diet fatigue and other negative side effects of being at a >500 calorie deficit may make the diet too difficult to maintain.
I actually went over this recently with someone who wanted to build something similar. The conclusion was this is a very difficult problem to solve, probably intractable to some extent. You can't see the complete composition of the food with a standard camera. E.g. I make a salad which is maybe 300 calories. Then I sprinkle some croutons and bacon on top, which will mostly be in the middle. Then I put dressing on it, which is hard to estimate. That dressing hides the bacon and croutons and, since it contains a lot of oil, could seriously skew the measurement one way or the other. Now I mix it all around and the AI can't tell how much dressing was used at all.
I pick this example because I've seen specifically this cause problems for people trying to lose weight. They think their eating a salad, not realizing they've thrown an extra 500 calories on top.
Another case: I sit down to breakfast, having made myself eggs and toast. One of if not the largest contributor to my calorie intake will be the amount of butter on my toast. If I use four pats that will probably exceed my calorie intake from eggs. If I use one, not as much. I sincerely doubt it's realistic to tell the difference with any sort of precision.
I'm in the calorie identification app business, and one benchmark I like to use is the old milkshake salad. It LOOKS like a salad with some mystery white sauce, maybe a ranch dressing or some such, but it's actually a milkshake. Hah! Gotcha. Humans: 3, AI:1.
I think even if you have volume factored in like photogrammetry/lidar, the inside of the food could be something else too like a chunk of meat inside potatoes
Anecdote of one... about 8 years ago I put a lot of effort into losing weight... when I was eatng a 1500 calorie/day diet, I wasn't losing weight... per a nutritionist friend who looked at what I was eating suggested closer to 2800 (6'1" tall, 370# at the time) and when I did, I started losing weight. Note this was combined with crossfit 3-4 days a week both on the lower and higher calorie intake.
If you have a really dysregulated metabolism, your body can definitely work against you when consuming too little.
Had a friend who came to me with the same idea. In challenge of the idea I sent a picture of some chopped cucumbers and tomatoes, and an other picture with the same chopped vegetables drizzled with a copious amount of olive oil. I asked if they can tell which one is which. The difference was hundreds of calories and you could not tell which one is which (least of all how much oil there is on the one with the oil.)
That of course feels like a "weird" edge case, but it illustrates the general problem that butter/oil/sugars can pack a lot of calories and have no or almost no visual signature.
You're right, but I think the general public is largely ignorant about calories and nutrition. Even basics like 4 cals per gram of protein and carbohydrate, 7 for alcohol, 9 for fats. Or even what a carbohydrate really is versus a "sugar", which people talk about as bad.
There's a sweet spot for an app that is inaccurate with a market that wants it but doesn't understand how inaccurate it is.
Kind of like how I could vibe code an app, get it to "work", think it's great and be ignorant of the many ways it will break or isn't working that a knowledgeable developer could.
The general public's understanding of food is still stuck in 1990s "destroy all fats (and completely ignore the sugar we replaced it with)" diet culture.
The fact that people still believe in reducing fat as it's own goal (instead of being an easy way to reduce calorie content) is a testament to how bad the public is at identifying fact from fiction.
Then you have shit like the influencer foods, "Feastables" and "Hydration beverage" Prime, which is just flat soda. It's pathetic.
Or think of all the dude bros who insist on dry scooping cup fulls of protein+caffeine powder, and going home to gorge themselves on two pounds of chicken breast, and yet doing absurdly normal amounts of weightlifting or exercise that requires no modifications to their already protein overloaded American diet.
Diet culture is what is fucking American health. People read fucking tabloids that bad-faith regurgitate poorly done "science", funded by the council for selling more food, and insist that since "Woman's Health" says that scientists say chocolate both kills and saves you, scientists are dumb and know nothing, even though THE ACTUAL SCIENCE NEVER CONCLUDED ANYTHING, because the scientific paper was just an observational study!
Everyone keeps using this example, but you have the exact same problem calorie counting any food that might be lathered in fat. It's why people underestimate their calorie intake by 30%+.
You could level the same criticism at Cronometer and MacroFactor when you try to log food you received at a restaurant. Yet those apps are still useful (and I think requisite) for knowing what you're eating. And you should probably 1.5x the calorie estimation when you eat out.
What's interesting is whether this app can accurately estimate food at all. If it can, then that's a huge win and you can add your own buffer zone for oils like you already have to do when you count calories. ...Or chill on the butter and restaurant food when you're supposedly trying to lose weight.
> Everyone keeps using this example, but you have the exact same problem calorie counting any food that might be lathered in fat.
Not really. In practice you need to know the ingredients to estimate the caloric value. Either because you prepared the meal, or because someone who wrote the recipe of it calculated and wrote it on the packaging/menu.
> If it can, then that's a huge win
But that's the point of the example. That it can't. If it could, that would be good. But it can't do it, and not because the app is deficient in some way, but because the necessary information is not available in the image.
> Or chill on the butter and restaurant food when you're supposedly trying to lose weight.
Yes of course. And that is one of the things you learn when you do calorie counting. The practice drives home that message, and many others. But you are not going to learn that if the app hides the signal from you.
The 1st issue I'd think of would be scale.. eg. identical looking plate of food in EU vs USA. The images might look identical but the USA one would be scaled up and contain much more calories.
Everything that could be in the image for scale reference would also be scaled up along with the food.
My reasoning for evaluating the relevancy of this app:
1. If the food being scanned has a nutrition label, I don't need this app.
2. If there's no nutrition label, the app can't possibly extrapolate the nutritional information. It can't estimate portion size or the ingredients in the recipe.
3. If the app can't extrapolate nutritional information, I don't need this app.
I think the key insight here is that actually counting calories is not needed by the app's demographic. Possibly just getting them to pay attention to what they are eating is enough for them to see progress or at least feel like they are doing something.
Recently I have started running dish descriptions through deepseek-r1:8b to give a rough estimate on calories and it is sufficient, at least for my cooking. Of course this will hallucinate on complex cases, and arguably simple dishes are easy to count. Still, counting by hand is tedious compared.
I am happy with a rough number on a dish, I do not need precision to know if I am not eating enough calories. Note there are errors on the result and reasoning, but still this dish should be around the ballpark of 500 cal. I figure running this through a larger model might be more accurate. I wondered if maybe I wanted to create an app for that, but I am happy not to.
Yes, simple descriptions could be parsed with NLP, and yes, calories could be inferred with a database.
>>> Calorie count of a dish that has: rice, broccoli, two pieces of turkey, and a fri
... ed egg. Assume this is a normal portion, for instance, there are 80g of rice.. Re
... turn the result in a json structure
<think>
Alright, so I need to figure out the calorie count for a dish that includes
rice, broccoli, two pieces of turkey, and a fried egg. The user mentioned it's
a normal portion, like 80g of rice. Hmm, okay, let me break this down step by
step.
First, I should find out how many calories are in each ingredient separately.
Let's start with the rice. Rice is a common staple, so I think 80g of cooked
rice has about 130 calories. I've seen that number before, but I should
double-check to make sure it's accurate.
Next up is broccoli. Broccoli is pretty low in calories, right? A portion of
around 100g should be about 25-30 calories. Wait, maybe a bit more if it's raw
or cooked differently, but I think for this purpose, 35 calories is reasonable.
Then there are two pieces of turkey. Turkey can vary based on the cut and
whether it's cooked. Let's assume these are regular chicken breast slices, each
around 50g. Each slice would have about 110 calories, so two pieces would be
around 220 calories. But I should make sure that I'm not confusing turkey with
another type of meat. Turkey does tend to be a bit higher in calories than
chicken sometimes, but 110 per slice seems right.
Lastly, the fried egg. A single fried egg is usually about 140-150 calories
depending on how it's made. Since there's just one egg, that's around 140
calories. I think that's correct because yolks add a good number of calories
due to fat content.
Now, adding all these up: rice (130) + broccoli (35) + turkey (220) + egg
(140). Let me sum them: 130 plus 35 is 165, plus 220 makes 385, and then plus
140 gives 525. So the total would be around 525 calories.
Wait a second, maybe I should consider if there are any other ingredients or
whether portions were different. For instance, sometimes turkey portions can be
larger or smaller depending on the meal. But since it's specified as two
pieces, each at about 50g, that seems okay.
Also, considering preparation methods: fried egg might have more calories if
it's done in a certain way, but I think one fried egg is standard. Broccoli's
calorie count could vary slightly, but for simplicity, the 35 seems fine.
Is there any chance I'm missing something? Maybe adding spices or sauces, but
unless specified, I shouldn't assume that. So the total should be around 525
calories in a normal portion as described.
</think>
```json
{
"calories": {
"rice": 130,
"broccoli": 35,
"turkey": 220,
"fried egg": 140,
"total": 525
}
}
```
Most restaurants smuggle obscene amounts of butter and sugar into dishes to make them more delicious so you'll come back. It's not unreasonable to assume that every subcomponent of a recipe might have sugar added individually.
In packaged foods, there is a whole science of masking the sugar and fat content to make it more addictive without triggering your inbuilt satiety mechanisms [1]. This is what today's engagement optimisers did for money in the 50/60s.
You could argue that these "innovations" were precisely to subvert the intuition that visual appearance of food (and other natural sensors) can be relied on to assess their nutritional properties.
There is absolutely, unequivocally, 0 chance this can be accurate within any kind of reasonable bounds. I'm guessing you haven't done much calorie tracking if you think this could possibly be true.
I can make two dishes that look identical and have +/- 50% caloric content, easily.
> I expect a strong correlation between the visual appearance of food and its caloric content
This doesn't pass even simple scrutiny. There are so many caloric ingredients that aren't visible in food. You can't tell just by looking whether a rice dish contains half a stick of butter.
This is completely wrong. For example, you can increase the amount of oil or butter in a recipe, doubling or tripling its calorie count, and you would never be able to tell from a picture.
Many people will analyze this from a tech perspective. I urge you not to. This isn't about whether the technology is good or bad. The real reason they succeeded is their mastery of marketing, particularly TikTok and influencer marketing. Understanding distribution is far more important than knowing how to build something, and this is truer now than ever before. I know it stings but its the truth.
I didn't analyze this for either: rahter what is what is the typical outcome for young humans who achieve a high degree of success/fame before their prefrontal cortex is fully formed.
As we understand more about brain development in kids, I think and under studied aspect are kids who got access to a lot more money than normal typical kids have, and the results such wealth brings.
> I think and under studied aspect are kids who got access to a lot more money than normal typical kids have, and the results such wealth brings.
As opposed to the kids born on planets without atmospheric oxygen, of course. Those kids don't stand a chance.
There's always some advantage others have. Money, good looks, distribution, connections, right place right time.
"Wealth inequality" isn't going to disappear overnight, and lamenting about it won't get us closer to success.
We're all dropped into the Darwinian gradient landscape. Some of us have better starting positions. While we sit and wait for policy decisions to make things easier, our job remains to find gradients that aren't too steep, aren't over-explored by the masses, and that give us some modicum of joy to spend our lives upon.
The minnows and gazelles have it much worse than us. Praying mantises, anglerfish, and the hosts for the parasitoid wasps are practically living in a daily Kafkaesque horror. Meanwhile we're in our own dopamine drip Disneyland with near-infinite degrees of freedom and plenty of years on the clock (for most of us).
The point of looking at this from a technical perspective though is that the underlying tech doesn't actually work, and it's just a smoke and mirrors game built on marketing.
Like Theranos. And Nikola. And Fyre Festival. Etc.
Age is irrelevant, unless we're trying to tell youngsters that fraud is an acceptable means to get ahead in life. But then again given our current political environment maybe that is what we're trying to tell the next generation.
EDIT: the app now also just reads nutrition labels as a backstop. Nutrition labels already include the calorie data, so the app isn't doing anything there.
> But then again given our current political environment maybe that is what we're trying to tell the next generation.
There's no trying.
The US has played "Just grift your way to money" as a standard means of operating since at least the dotcom boom. A reason so many young and poorly educated people jump at obvious MLM type scams and other scams is because they feel that everything is a scam, so surely they can get in on it too, right?
And everything IS a scam. Coke doubled in price in the past 5 years. I promise you their costs did not double. Their costs are Labor (highly automated), water (they almost always have sweetheart deals for dirt cheap water, cheaper than you pay), and one of the most subsidized commodities available, HFCS, or alternatively, a sprinkle of dirt cheap chemicals for their diet sodas.
People feel that, even when they don't understand or even recognize it.
People recognize that the US has been a scammy free for all for decades now. Everyone for themselves, fuck you, got mine.
We are on like the third generation raised this way. The people who took "Greed is good" to heart had kids, and raised them with it as a core principle.
Jordan Belfort, the guy who Wolf Of Wall Street is based on and spent time in prison for scamming his clients in basically the same way modern crypto pump and dumps work, now sells out auditoriums as a motivational speaker for fucks sake.
That's unlikely. Try driving in a snowstorm, where visual inputs become effectively useless, and you quickly realize how much the motion inputs are factored in as well.
They are very different problems. You can attach every kind of sensor to the phone and get HD images, lidar scan, 3D model, heat signature and whatever other signal you want, but you still won't be able to figure out calories in a piece of food without breaking it apart at a molecular level. There's just too much hidden information. What sensor will tell you how many tablespoons of oil the chef used to fry onions for the sauce? Heck the chef himself probably won't be able to tell you.
I legitimately drove full month with near zero interventions in FSD. The second statement is achievable in my opinion. Most people have not experienced hw4 with latest releases.
>But he looked around. “We were surrounded by people that were in their late 20s or 30s all day. And I realized that if I didn’t go to college, this is what life would be like.”
Lol, I can relate. I started working in an office when I was 16, now 24, and regretting wasting my youth grinding when I could have been having fun in a period of your life you only get to experience once.
Don't grow up too fast kids. Make stupid decisions and ride out your youth as long as you can afford to.
You're 24? Don't take this the wrong way but you have loads of time to do "youth shit."
As you get older, people expect you to be more competent in life and work. If you leave work for two years to bike around the world, it'll be a fabulous adventure and will in the grand scheme of things have little consequence down the road. Try that when you have a kid!
You're just defining your own fate here if you think 24 is so much older and different than 16. But also, perhaps your situation now would be different without the grind you went through? Instead of regretting anything I think you could just try to enjoy your youth now.
Meeting people is 10x harder when you’re an adult. Missing the boat on social groups in my younger years has made it extremely difficult to meet people and establish a social circle. Mid twenties is when people tend to pair up long term as well, and you’re not really going to see friends often when they have a spouse. Mid twenties is absolutely different to late teens.
>>But he looked around. “We were surrounded by people that were in their late 20s or 30s all day. And I realized that if I didn’t go to college, this is what life would be like.”
Its just a marketing stunt. They complain about not getting into Harvard but he could literally rent a penthouse in Cambridge, hang out on the campus throw big parties and get 90% of the experience while supposedly running a 30M ARR company. Absolutely no benefit to actually enrolling except ego.
I understand the scepticism here. For sure this app isn’t 90% accurate in any traditional sense.
One note, as someone who also built a calorie tracking app with ai as well as lost a good amount of weight with it: accuracy for calorie tracking doesn’t matter. You can honestly just detect if it’s a meal and return 600 cals. For most people the simple fact that they become aware of what they eat and think about their food in an ongoing basis will lead them to loose Weight.
Sticking to it is the hard part.
It matters a lot if you have specific macro goals. If you don’t want to lose muscle, it is important to eat high protein especially during a cut. And keeping fat and/or carbs low while doing this is quite difficult without accurate data.
> You can honestly just detect if it’s a meal and return 600 cals.
That's ridiculous. What signal would that provide to the user? Let's say someone who is eating double the portions they should be eating. How will this hypothetical app help them figure that out?
> For most people the simple fact that they become aware of what they eat and think about their food in an ongoing basis will lead them to loose Weight.
And you think people will " become aware of what they eat" by shooting a picture of their food with an app which always say "600 cal"? I don't think you thought this through.
Readit News
I'm all for making it easier for people to lose weight but this app may honestly have the reverse effect. If the app estimates calories too low (and therefore the individual eats more), many people will get frustrated with the lack of progress and give up. If the app estimates too high, the individual will lose weight, but diet fatigue and other negative side effects of being at a >500 calorie deficit may make the diet too difficult to maintain.
I pick this example because I've seen specifically this cause problems for people trying to lose weight. They think their eating a salad, not realizing they've thrown an extra 500 calories on top.
Another case: I sit down to breakfast, having made myself eggs and toast. One of if not the largest contributor to my calorie intake will be the amount of butter on my toast. If I use four pats that will probably exceed my calorie intake from eggs. If I use one, not as much. I sincerely doubt it's realistic to tell the difference with any sort of precision.
https://www.youtube.com/watch?v=fXWq2AVlrsY
Any other interpretive situation based solely on a camera has so many inherent flaws as to render this almost useless.
If you have a really dysregulated metabolism, your body can definitely work against you when consuming too little.
That of course feels like a "weird" edge case, but it illustrates the general problem that butter/oil/sugars can pack a lot of calories and have no or almost no visual signature.
A salad wouldn't even be the hardest case to detect, since raw vegetables don't soak up as much oil as other kinds of food.
There's a sweet spot for an app that is inaccurate with a market that wants it but doesn't understand how inaccurate it is.
Kind of like how I could vibe code an app, get it to "work", think it's great and be ignorant of the many ways it will break or isn't working that a knowledgeable developer could.
The fact that people still believe in reducing fat as it's own goal (instead of being an easy way to reduce calorie content) is a testament to how bad the public is at identifying fact from fiction.
Then you have shit like the influencer foods, "Feastables" and "Hydration beverage" Prime, which is just flat soda. It's pathetic.
Or think of all the dude bros who insist on dry scooping cup fulls of protein+caffeine powder, and going home to gorge themselves on two pounds of chicken breast, and yet doing absurdly normal amounts of weightlifting or exercise that requires no modifications to their already protein overloaded American diet.
Diet culture is what is fucking American health. People read fucking tabloids that bad-faith regurgitate poorly done "science", funded by the council for selling more food, and insist that since "Woman's Health" says that scientists say chocolate both kills and saves you, scientists are dumb and know nothing, even though THE ACTUAL SCIENCE NEVER CONCLUDED ANYTHING, because the scientific paper was just an observational study!
You could level the same criticism at Cronometer and MacroFactor when you try to log food you received at a restaurant. Yet those apps are still useful (and I think requisite) for knowing what you're eating. And you should probably 1.5x the calorie estimation when you eat out.
What's interesting is whether this app can accurately estimate food at all. If it can, then that's a huge win and you can add your own buffer zone for oils like you already have to do when you count calories. ...Or chill on the butter and restaurant food when you're supposedly trying to lose weight.
Not really. In practice you need to know the ingredients to estimate the caloric value. Either because you prepared the meal, or because someone who wrote the recipe of it calculated and wrote it on the packaging/menu.
> If it can, then that's a huge win
But that's the point of the example. That it can't. If it could, that would be good. But it can't do it, and not because the app is deficient in some way, but because the necessary information is not available in the image.
> Or chill on the butter and restaurant food when you're supposedly trying to lose weight.
Yes of course. And that is one of the things you learn when you do calorie counting. The practice drives home that message, and many others. But you are not going to learn that if the app hides the signal from you.
Spoiler: It can't. It is physically impossible to determine calories from pictures of food.
This was technically correct but missed out on a viral app and millions of revenue?
1. If the food being scanned has a nutrition label, I don't need this app.
2. If there's no nutrition label, the app can't possibly extrapolate the nutritional information. It can't estimate portion size or the ingredients in the recipe.
3. If the app can't extrapolate nutritional information, I don't need this app.
4. I don't need this app.
What you need is a LLM.
Deleted Comment
I am happy with a rough number on a dish, I do not need precision to know if I am not eating enough calories. Note there are errors on the result and reasoning, but still this dish should be around the ballpark of 500 cal. I figure running this through a larger model might be more accurate. I wondered if maybe I wanted to create an app for that, but I am happy not to.
Yes, simple descriptions could be parsed with NLP, and yes, calories could be inferred with a database.
Deleted Comment
Dead Comment
In packaged foods, there is a whole science of masking the sugar and fat content to make it more addictive without triggering your inbuilt satiety mechanisms [1]. This is what today's engagement optimisers did for money in the 50/60s.
You could argue that these "innovations" were precisely to subvert the intuition that visual appearance of food (and other natural sensors) can be relied on to assess their nutritional properties.
[1] https://en.m.wikipedia.org/wiki/Bliss_point_(food)#:~:text=T...
I can make two dishes that look identical and have +/- 50% caloric content, easily.
This doesn't pass even simple scrutiny. There are so many caloric ingredients that aren't visible in food. You can't tell just by looking whether a rice dish contains half a stick of butter.
1 tbsp of olive oil has 135 calories.
How would the app know which fat the food was cooked in?
As we understand more about brain development in kids, I think and under studied aspect are kids who got access to a lot more money than normal typical kids have, and the results such wealth brings.
As opposed to the kids born on planets without atmospheric oxygen, of course. Those kids don't stand a chance.
There's always some advantage others have. Money, good looks, distribution, connections, right place right time.
"Wealth inequality" isn't going to disappear overnight, and lamenting about it won't get us closer to success.
We're all dropped into the Darwinian gradient landscape. Some of us have better starting positions. While we sit and wait for policy decisions to make things easier, our job remains to find gradients that aren't too steep, aren't over-explored by the masses, and that give us some modicum of joy to spend our lives upon.
The minnows and gazelles have it much worse than us. Praying mantises, anglerfish, and the hosts for the parasitoid wasps are practically living in a daily Kafkaesque horror. Meanwhile we're in our own dopamine drip Disneyland with near-infinite degrees of freedom and plenty of years on the clock (for most of us).
Seneca said some good things about this.
Like Theranos. And Nikola. And Fyre Festival. Etc.
Age is irrelevant, unless we're trying to tell youngsters that fraud is an acceptable means to get ahead in life. But then again given our current political environment maybe that is what we're trying to tell the next generation.
EDIT: the app now also just reads nutrition labels as a backstop. Nutrition labels already include the calorie data, so the app isn't doing anything there.
There's no trying.
The US has played "Just grift your way to money" as a standard means of operating since at least the dotcom boom. A reason so many young and poorly educated people jump at obvious MLM type scams and other scams is because they feel that everything is a scam, so surely they can get in on it too, right?
And everything IS a scam. Coke doubled in price in the past 5 years. I promise you their costs did not double. Their costs are Labor (highly automated), water (they almost always have sweetheart deals for dirt cheap water, cheaper than you pay), and one of the most subsidized commodities available, HFCS, or alternatively, a sprinkle of dirt cheap chemicals for their diet sodas.
People feel that, even when they don't understand or even recognize it.
People recognize that the US has been a scammy free for all for decades now. Everyone for themselves, fuck you, got mine.
We are on like the third generation raised this way. The people who took "Greed is good" to heart had kids, and raised them with it as a core principle.
Jordan Belfort, the guy who Wolf Of Wall Street is based on and spent time in prison for scamming his clients in basically the same way modern crypto pump and dumps work, now sells out auditoriums as a motivational speaker for fucks sake.
The kids LOOK UP TO SCAMMERS
“You can have a full self driving car with just a few cameras”
In a way both things are very much similar and the real accuracy is more of a fiction than reality.
Deleted Comment
That's unlikely. Try driving in a snowstorm, where visual inputs become effectively useless, and you quickly realize how much the motion inputs are factored in as well.
Lol, I can relate. I started working in an office when I was 16, now 24, and regretting wasting my youth grinding when I could have been having fun in a period of your life you only get to experience once.
Don't grow up too fast kids. Make stupid decisions and ride out your youth as long as you can afford to.
As you get older, people expect you to be more competent in life and work. If you leave work for two years to bike around the world, it'll be a fabulous adventure and will in the grand scheme of things have little consequence down the road. Try that when you have a kid!
Its just a marketing stunt. They complain about not getting into Harvard but he could literally rent a penthouse in Cambridge, hang out on the campus throw big parties and get 90% of the experience while supposedly running a 30M ARR company. Absolutely no benefit to actually enrolling except ego.
One note, as someone who also built a calorie tracking app with ai as well as lost a good amount of weight with it: accuracy for calorie tracking doesn’t matter. You can honestly just detect if it’s a meal and return 600 cals. For most people the simple fact that they become aware of what they eat and think about their food in an ongoing basis will lead them to loose Weight. Sticking to it is the hard part.
That's ridiculous. What signal would that provide to the user? Let's say someone who is eating double the portions they should be eating. How will this hypothetical app help them figure that out?
> For most people the simple fact that they become aware of what they eat and think about their food in an ongoing basis will lead them to loose Weight.
And you think people will " become aware of what they eat" by shooting a picture of their food with an app which always say "600 cal"? I don't think you thought this through.
Take a picture of everything you eat and correlate it with symptoms. Have AI figure out what may be a trigger.
(I have a super rare food disease that took years to figure out and made my life unbearable).