Sunday, August 05, 2018

Food scanner scams: non-physicists overestimate the "diversity of food"

They also underestimate the importance of expertise, experience

Hours ago, Thunderf00t posted a wonderful new 50-minute-long video:

ThunderF00t chooses "Pražská vodka", i.e. the "Vodka of Prague", as his template for an ethanol-water mixture. He deserves special compliments for that. That Czech vodka is based on the 1978 diploma thesis of Ing. Eugen Skalický at Prague's University of Chemical Technology and uses the Highland (Vysočina) Wheat to produce the ethanol. It's been Czechia's bestselling vodka, is still near the top, but I have no idea how he got one unless he is just visiting my homeland.

It shows how utterly useless two smartphone-based, StarTrek-based food scanners or spectroscopes are – and why their crowdfunding campaigns that had collected $3 million and $0.4 million must be considered scams.

These two bogus revolutionary products, TellSpec and Scio, cost $149 or $250 (plus $1,000 for software etc.), respectively. You point them to a food, some photons are reflected, the spectrum is evaluated, and you know what's inside the foor or pills perfectly, the nutrition values, and everything else.

Needless to say, it's bullšit. You can't get more information from the edible stuff than what the existing spectrometers do. And the existing spectrometers cost over $1,000 and $70,000, respectively, and even they can't fulfill the promises for the cheap products. You immediately face problems because the apples and oranges are not transparent for the infrared light. The light gets absorbed by the skin of the orange. If you peel it, the light gets scattered in all directions, and so on. If enough light doesn't get back to the spectrometer, your signal will be weak or non-existent.

If you get inside the food, you may see some characteristic frequencies of the sugar, but the spectrum of almost everything else looks dull and rather featureless (and most features you can see are really "noise"). You can't even distinguish apples from oranges or pears, the king of cheese from the cheapest cheese, sometimes even apples from tomatoes. On the other hand, the clear differences in the spectrum arise from rather nutrient-wise irrelevant differences e.g. between the yellow, green, and red skin of an apple. There's really another problem he doesn't discuss that the different components of the food have different albedo so the ratios of the components can't be trusted.

If you want to to identify the pesticides in the fruits and other things they promised, you would have to fight with the tiny concentrations and the low sensitivity of the gadgets. OK, so the gadgets for thousands of dollars – or $70,000 for a Raman spectrometer (wavelength of 10 microns plus minus 30%) that Thunderf00t got for $6,000; and a near-infrared (close to 1 micron) which may cost as little as $1,000 – can do much more than these toys, but it's still completely useless for getting any accurate enough information about the food that has many components: You can see at most the surface and different materials that appear in the food look almost identical up to the unavoidable noise and error margins, anyway.

Moreover, if you wanted to use these gadgets for gain loss, it's utterly pointless. If you want to lose weight, there is a magic recipe: Just eat less food. This is a miraculous recommendation that none of the people obsessed with the diets has discovered yet. But it's the only one that really works. Even if gadgets similar to Scio and TellSpec could measure the precise percentage of various nutrients (be sure that the nutritional labels give you more helpful information!), you wouldn't know how much you're eating. Different parts of your hamburger have different percentages of carbohydrates, fats, and proteins, and other things and you can't really measure the weight and percentages in each part separately, especially when the parts are continuously switching from one to another. Imagine how complicated the hamburgers or salads are in McDonald's – and almost all other food is even more complicated.

One part of our mass culture is the obsession with food quality. Lots of people believe in some truly supernatural abilities and health hiding in the divinely right proportion of the nutrients etc. Most of this stuff is bullšit as well, of course. You need some energy – which can be obtained from carbohydrates, sugars, and even proteins, and those can be largely converted to each other (and the detailed choice which kind of a sugar etc. is almost irrelevant) – plus some dietary fibers, minerals, and vitamins... I may have forgotten one or two similar classes of compounds but that's it.

(And it's probably even more important to avoid really toxic or carcinogenic compounds in your diet – but I am really talking about some obviously harmful compounds, like pesticides, plutonium, or arsenic LOL, not compounds that are still just speculated as being slightly harmful.)

Unless you omit some essential nutrient, you must be OK. In fact, the body finds the precise composition so irrelevant that even those people who deliberately omit some really vital and major parts of their diet – e.g. vegetarians and vegans – may still live rather healthy lives although I am pretty sure that the inferiority of their diet already has observable consequences (it has some advantages, too). But the subtler differences between the types of products are almost completely irrelevant for one's physical well-being.

Lots of people share my "indifferent" view. I am willing to claim that this is the view of every real competent scientist. But there are lots of people who are really obsessed with watching what is the newest recommendation by the "food experts" what to eat and what to avoid. Well, my father is a textbook example. Now for almost 70 years, he's been obsessed with seeing every single fudging article in the Czech media about these matters. Roughly once a week, he would radically change his opinions about a critical question. Are saturated fats healthy or not? Most recently, do Omega-3 fatty acids cause cancer or prevent cancer (e.g. prostate cancer)? Over the years, the journalists have written lots of contradictory press about those – and indeed, there are contradictory papers in the scientific literature, too.

If these claims are changing so often, it proves that there's some major uncertainty, the latest article probably isn't the last one, and one shouldn't be obsessed with these reports, anyway, should he? To me, this looks like an absolutely rudimentary basis of any rational thinking about these matters. But for some reasons, my father can't possibly "agree" with that. And he's not the only one. I think that lots of people are obsessed with such things.

I think that this kind of obsession with some miraculous advantages of the "optimal composition of food" is a widespread pseudoscience that is largely driven by the people's conflation of their taste sensors and rational thinking about the health of the food. Our taste sensors can distinguish a rather large number of types of food. It's intriguing for many people to imagine that there's some holy point in this parameter space that tells you how to live for 120 years and approach God.

(Incidentally, I wrote 120 years because Bogdanov Brothers' Sister Veronique – I wrote a French book about theoretical physics and about them, L'Equation Bogdanov – has claimed that the brothers' increasingly extraterrestrial appearance is a result of their high-tech modifications of their own bodies that will allow them to live for 120 years. Cool!)

Well, I don't think you can really surpass 120 years of age by adjusting the optimal number and types of apples and other fruits that you eat. The oldest people in the world have had very different diets, after all. So the very belief that there is some "magic precise truth about the diet" that you should find is an irrational religion of a sort. And after all, as one of the comedians from the U.S. cable TVs pointed out, most of the authors of books about the very healthy food have died significantly younger than the average life expectancy. What an irony. Does it tell you something?

Well, I think that this piece of data does tell something to everyone who is being rational. It's just a waste of time.

Spectrometers or spectroscopes may measure the intensities at any – continuous – frequency, i.e. infinitely many numbers that describe the spectrum. But in practice, there is only a "low-dimensional parameter space" of the possible spectra of the things that we eat. Moreover, most of these dimensions that describe the "food's spectrum" tell us very little about what kind of a food or nutrient it is – let alone about the overall health of the given food.

Because the spectrometers only see the surface of solids, they can't even distinguish M&Ms from some pretty special pills because the pills are often coated in the same sugar as M&Ms. I was surprised that some multivitamins are coated by titanium dioxide. ;-) There are lots of instinctive reactions – I've never studied it and it seems strange to eat titanium dioxide. But the folks who chose it probably had good reasons and checked it's safe enough. It also looks like a terrible idea to eat lithium, even lick micrograms of lithium that is left somewhere. But you may actually eat a gram of lithium per day and do pretty well. So the number of toxic things you really have to be careful about is rather limited. Well, pesticides, plutonium, and arsenic are bad.

A reasonably good spectroscopy expert could tell the science journalists very quickly that – and why – Scio and TellSpec are almost certainly bullšit. But none of the science journalists ever bothered to ask an actual scientist. So there's been a huge amount of hype about these products. Check e.g. this junk in the Washington Post (2016). The title said "A groundbreaking technology will soon let us see [inside food?]" and it said "I have seen the future of food transparency, and it is optical. Also, it fits in your smartphone. Imagine a scanner the size of a grain of rice, built...". Why couldn't they ask someone who could do the analysis and figure out that it's impossible to get much helpful information about the food from a similar gadget, especially from such a trimmed down version of a spectral analyzer?

And it's not just far left scum at publications such as WaPo. Check this utterly uncritical video at Bloomberg.

Aside from the "divinely healthy food" religion, I think that there exist additional postmodern religions that abuse some people's desire to look for the optimal point – while underestimating all the error margins and overestimating the importance of the differences. I think that the main ultimate consequence of most of these religions is that lots of wealthy enough people waste a lot of money for things that are expensive and hyped to be remarkable in some way – but they end up being the same as the cheap stuff bought by the poorer people. For example, lots of wealthy people waste money for the organic food even though most of the food sold as "organic" really isn't and on top of that, there's no reason to think that the "organic" food is healthier than the "inorganic" food.

I always loved this episode of Penn&Teller's Bullšit – all the episodes are wonderful – as a proof how the smug people behave irrationally. First, they discuss that bottled water is usually less good than tap water in the U.S., due to its lousy regulation by FDA. Then they conquered a very expensive Californian restaurant and through their bogus water steward, they served amazing expensive waters with fancy French names. The clients were impressing each other with the description of the special feelings they had while drinking that expensive water. But the source of all this water was a garden hose! ;-)

Many of us were educated so that we're not picky. I may have been picky when I was 5 but I am not picky anymore. Moreover, over the years, I found it extremely important to only care about the aspects of the taste etc. that I can actually distinguish and I sufficiently care about, not about some bogus taste characteristics that others tell me that I should care about – or others tell me that they can distinguish or find important. This approach seems like 1) a matter of scientific integrity to me, 2) a protection against overpaying for nonsense like the water from the garden hose.

There are various other generalizations of the Scio and TellSpec scams. Much more generally, lots of people buy the stories that some randomly built group of friends can suddenly build vastly cheaper and superior scientific instruments than what the professional scientists were capable of building so far, after the centuries of practice. In principle, it's possible. Sometimes, it may happen. But a person who has at least the basic scientific literacy must know that it's very unlikely and it occurs very, very rarely.

One reason why so many laymen buy this stuff is their postmodern religious search for the "optimal point" in situations where no optimal point really exists or where the choice of the point isn't too important. Another reason is their distrust in professional scientists, including the very experimental ones who work with spectroscopes and similar gadgets all the time, combined with the wishful thinking that "ordinary people similar to themselves trump the scientists in science". These laymen vastly overestimate the probability that an outsider makes (or even can make) a revolution. If they studied at least a few similar cases and what was the resolution, they would have to understand that almost all such claims – automatic generators of water from the air (dehumidifiers that are claimed to be better than all others, also debunked by ThunderF00t), Scio and TellSpec, Andrea Rossi's cold fusion, and hundreds of similar examples – are almost guaranteed to be rubbish.

A sociological reason why they're almost guaranteed to be rubbish is that the outsiders who claim to become leaders in these gadgets just can't possibly have the skills or special talents to get to the top. In sports or music, most people would understand that someone who doesn't look like the top athletes or musicians – and who has gone through no similar stages of his life – probably can't be the leader. But for these laymen, spectroscopy or any other part of physics or natural science must be "less meritocratic than sports and music", and therefore the experience and expertise doesn't matter according to them.

But it does matter. It matters much more than it does in sports and music, in fact.

Another common sense consideration that should allow an intelligent layman to figure out that such things don't work is the realization that if they could work, someone would have already used them. As ThunderF00t also mentions, if the expensive spectrometers could be made outdated by some smartphone-like cheap device, the replacement would have probably taken place already. It hasn't taken place yet so it's probably not so easy. The food scanners are just an example of this reasoning. The same applies to cold fusion or anything else. Great labs surely have better equipment and experts to do most things that Andrea Rossi claimed to do, so if there were something truly interesting in his work, they would have done it before – or shortly after him.

But lots of the laymen don't "agree" with me – more precisely, they don't understand this simple reasoning. A part of it is that they don't want to understand it. The wishful thinking is a part of their psychological well-being – it's a contemporary replacement for the religion. And people believe lots of such postmodern religions, indeed.

I think that the crowdfunding campaigns should be equipped with some mechanisms that protect the donors' money in the case of the scams of the type that "the proposed gadget with the promised functionalities and characteristics is physically impossible", like this case. At least the people who want to be told about some scientific views should be able to get them. And scientists like ThunderF00t should at least have some power to issue official verdicts on whether or not the promised gadget was realized – or whether it was possible. Without that, the crowdfunding servers are almost guaranteed to be largely composed of similar scams.

No comments:

Post a Comment