Slashdot Mirror

← Back to Stories (view on slashdot.org)

Machine Learning Is Making Pesto Even More Delicious (technologyreview.com)

Posted by BeauHD on from the benefits-of-artificial-intelligence dept.

An anonymous reader quotes a report from MIT Technology Review: Machine learning has been used to create basil plants that are extra-delicious. While we sadly cannot report firsthand on the herb's taste, the effort reflects a broader trend that involves using data science and machine learning to improve agriculture. The researchers behind the AI-optimized basil used machine learning to determine the growing conditions that would maximize the concentration of the volatile compounds responsible for basil's flavor. The study appears in the journal PLOS One today.

The basil was grown in hydroponic units within modified shipping containers in Middleton, Massachusetts. Temperature, light, humidity, and other environmental factors inside the containers could be controlled automatically. The researchers tested the taste of the plants by looking for certain compounds using gas chromatography and mass spectrometry. And they fed the resulting data into machine-learning algorithms developed at MIT and a company called Cognizant. The research showed, counterintuitively, that exposing plants to light 24 hours a day generated the best taste. The research group plans to study how the technology might improve the disease-fighting capabilities of plants as well as how different flora may respond to the effects of climate change.

32 of 79 comments (clear)

  1. Machine Learning by Anonymous Coward · · Score: 1

    Can it grow romaine lettuce without the E.coli?

  2. sustainable? by Anonymous Coward · · Score: 2, Interesting

    Okay so to "revolutionize farming" what percent of crops need to be grown indoors under artificial light and what is the financial and environmental cost of that? I support progress and things like geothermal greenhouses if they reduce energy used in shipping. If you put a greenhouse partially in ground and use an air intake buried underground you can grow things that would be otherwise shipped in. Even if you supplement the lighting a bit that might be a win but I don't see how this revolutionizes agriculture in a sustainable way.

    https://www.usatoday.com/story...

    1. Re:sustainable? by alvinrod · · Score: 2

      If you've got a completely controlled environment, herbicide and pesticide become unnecessary, which would be an immense change. If you're also using other techniques to maximize the yield, nutrition, or other properties while doing that you can probably cut down on waste as well. There are some crops where some of it's left unharvested or just tossed if it doesn't meet some quality characteristics. Being able to localize food production is also something that's fairly important and helps provide jobs in those communities as well. People are pretty disconnected from where their food comes from these days, hence the meme of city kids thinking that meat comes from the grocery store.

    2. Re:sustainable? by del_diablo · · Score: 1

      Now there is a lot of interesting things here, for academia and research.
      But for long term real world usage you will eventually erode the soil, since there isn't a really really big ecosystem supporting it. So then you break the cycle to get more fertile soil. What is useful information will be how soil behaves in a closed system, and what data you can gather that will end up having useful real world applications.

      But there is also another use for this: A lot of food is tasteless when you accelerate its growth and then harvest it too early. Lets say herbs like Dill or Ruccola.
      Part of the reason why its tasteless is that there is a mismatch between the processes that happens, the day/night cycle plants experience, and how the nutrients are used.
      This can be said for all monoculture or industrial agriculture. Or even livestock production.

    3. Re:sustainable? by TheCastro1689 · · Score: 1

      This will work in those super towers they'll build in the future where people live and work and do everything inside a building they never leave.

    4. Re:sustainable? by Aighearach · · Score: 1

      The "revolutionary" aspect is that they did a farming experiment where they tried to maximize taste.

      Normally farming experiments only try to maximize yield, durability, and resistance to toxins.

    5. Re:sustainable? by Aighearach · · Score: 1

      It is a bit hard to stay alive if you leave the bubble. Are you sure you didn't just sneak in this morning? Wait right here while I have security scan your ID implant.

  3. Re:Hmm by Anonymous Coward · · Score: 2, Interesting

    I've tried 24 hour lighting before and it sucks ass. After a while, my plants started looking droopy like they're exhausted and needed sleep. They also don't grow very much at all when you give them non-stop sunlight. Growth seems to predominantly occur during dark hours, using energy that was stored up during daylight hours.

    I got the best results by tweaking on/off cycles such that I had two effective days/nights during one 24 hour period. I could grow plants this way significantly faster than anything grown outside in the sun. Think of plants as little solar battery powered machines. They store up energy during the day and then spend that energy growing new tissue at night. Like batteries, there is only a limited capacity for storing so much energy at once, so it makes sense to create shorter artificial days/nights to speed up growth, especially for small young plants that have very limited storage capacity.

    Basil that I grew indoors made huge leaves that tasted better to me than outdoor grown basil.

    Strawberries grown indoors, unfortunately, taste like watered down crap. Apparently, strawberries need a lot more time and environmental stressors (UV, hot/cold cycling, water limiting, etc) to develop high quality tasting berries.

  4. Re: Hmm by lsllll · · Score: 1

    Is doesn't matter if it's good for them or not. They're slave to us!

    --
    Is that a roll of dimes in your pocket or are you happy to see me?
  5. Basil by Dunbal · · Score: 2

    All the basil plants I ever buy end up dying. I have no idea what I'm doing wrong.

    --
    Seven puppies were harmed during the making of this post.
    1. Re:Basil by jemmyw · · Score: 1

      Same. I love oregano but its a weed and grows everywhere. I let it flower this year and the bees love it. Basil and thai basil both producing more than I can use. Basil, dill, thyme, sage, marjoram, tarragon, I started from seed and once they germinated planted in the garden and they required no real care except trimming for use.

      I've made pesto with basil, thats quite fun. Drying it isn't as fun, you have to plan ahead a little because the best result was height of summer. And then I was too lazy to go outside and get fresh stuff and used all the dried stuff up instead of saving it for winter.

    2. Re: Basil by sodul · · Score: 1

      like out the toilet?

    3. Re:Basil by Anonymous Coward · · Score: 1

      Cilantro is coriander leaves. Coriander is the plant. Coriander spice is usually only the ground seeds. Cilantro is a Spanish word, not a "yanks" idiosyncrasy and quite ignorant of you to assume so.

      All the herbs you listed are often treated as edible ground cover, so your result is no surprise. Others include lavender and thyme. The last is the most virulent to competitors.

    4. Re:Basil by religionofpeas · · Score: 1

      Start with rosemary, much easier to keep alive (plenty of sun, let it dry before watering). Then try thyme and oregano.

    5. Re:Basil by _merlin · · Score: 1

      Are they getting enough sun and water? Is the climate suitable? I found that on the north coast of NSW (high humidity, hot summers, mild winters, no frosts, high rainfall) I could grow basil effortlessly. The plants thrived and grew huge with minimal care. However in Melbourne (less humidity, less rainfall, colder winters, occasional frosts) the basil plants wouldn't grow anywhere near as big and required a lot of care or they'd die.

    6. Re:Basil by Krakadoom · · Score: 1

      Rosemary is not easier, unless you have very specific environmental conditions where you live. Ie. hot and dry mostly. Rosemary hates getting it's roots wet and will die quickly that way. Which applies to most temperate areas.

    7. Re:Basil by Aighearach · · Score: 1

      I love oregano but its a weed

      I love oregano, but my wife thought it was a weed and plucked it, so now I'm eating it dried from a plastic can.

      I'm just glad the thyme was too difficult to pull up.

      Luckily she's Thai so I can at least grow basil safely.

  6. Tastes better? by Anonymous Coward · · Score: 5, Insightful

    I saw nothing in the links that showed the plants were *tasted* to see if they were better. Taste is a blend and highly subjective. Run those plants by some professional chefs so we can read their responses.

    1. Re:Tastes better? by omnichad · · Score: 1

      They should have just said tastes "stronger." It would be less false, less controversial, and not all that compelling either.

    2. Re: Tastes better? by Aighearach · · Score: 2

      Typical âoeAIâ nonsense story

      Kids these days with their funny words. I don't even know what âoeAIâ means! And honestly, I'm not even sure how many syllables it has.

  7. Strange tasting process by sad_ · · Score: 4, Interesting

    "The researchers tested the taste of the plants by looking for certain compounds using gas chromatography and mass spectrometry."

    instead of really tasting it? it only should taste better theoretically, there aren't any real taste test results that confirm the actual taste is actually better (even though, this is a very subjective thing).

    --
    On a long enough timeline, the survival rate for everyone drops to zero.
    1. Re:Strange tasting process by Ol+Olsoc · · Score: 1

      "The researchers tested the taste of the plants by looking for certain compounds using gas chromatography and mass spectrometry."

      instead of really tasting it? it only should taste better theoretically, there aren't any real taste test results that confirm the actual taste is actually better (even though, this is a very subjective thing).

      Yeah, looks like we have some bullshit going on here. Just because there is more of a compound that has been determined to taste good doesn't mean that it wouldn't be overbearing in larger amounts, or a different ratio to other compounds.

      Taste is subjective, but we gotta admit that someone has to taste something to tell what the taste is.

      Here's a cooking tip:

      If you are wilting cilantro in a saucepan, add a little coconut oil to it. This makes it easier to scrape that crap into the trashcan.

      This also works for kale.

      --
      The shepherds did so well protecting the flock that the sheep no longer believed that wolves existed.
    2. Re:Strange tasting process by Can'tNot · · Score: 1

      Think about that for a minute. Machine learning requires a tremendous number of iterations, with only very small changes between each one. You could do this with human tasters, but in order to measure those small changes accurately you would need a huge number of tasters and plants in order to get your error down to something reasonable.

      By first identifying which chemicals tasters want, and then measure those through far more precise means, you can go through those iterations much faster and more efficiently. And then presumably you verify with tasters every once in a while that you're headed in the right direction.

  8. Can it make pasta even more... by cemysce · · Score: 1

    AI-dente?

    1. Re:Can it make pasta even more... by Aighearach · · Score: 1

      No. Corn starch.

  9. Just wait by jfdavis668 · · Score: 1

    Soon the pesto will rise up and takes over.

  10. Re: Pesto? Disgusting but fashionable. by illiac_1962 · · Score: 1

    I grow basil specifically to make pesto.

  11. Great big DUH by AnalogDiehard · · Score: 1

    We needed a special research team and all that computer equipment to tell us what gardeners have known for DECADES? CENTURIES?

    When I buy a potted basil plant, there is a tag encouraging plenty of light. Gardening books spell out the care and feeding of all kinds of plants.

    Seriously, ALL THAT MONEY AND TIME was spent to do nothing more than reaffirm what we already know. Great big DUH!

    --
    Eternity: will that be smoking, or non-smoking? I Corinthians 6:9-10
  12. Genomic selection by hAckz0r · · Score: 1

    Genetic selection with reanimation, rather than just growing conditions, should be worked on as well. Once a plant is sitting on your dinner plate its evolutionary future is very bleak. That plant, no matter how wonderful it tastes to you, is simply an instance in history. A one-off thing of the past. If however, the genomic material could be sampled, saved, and reanimated in the form of a seedling, then the food industry could transform itself from serving the current tasteless cardboard, where every plant is a genomic clone of a single fungal disease resistant plant.

    Ever wondered what happened to the best-tasting bananas? Basically, they bread the taste out by breeding fungus resistance in. The growers and marketers let the taste take second place over 'production and packaging'. Once they find a resistant strain that packs and ships well (thick skin) they genetically remove the seeds, clone it in large numbers, and every plant on the plantation becomes an identical-twin seedling. Being able to select even the minor variations in taste, and positively selecting for it, could bring the natural taste back to the gastronomic marketplace.

  13. Ceaseless ignorance spreading... by argStyopa · · Score: 1

    First, every semi-automatic gun = machine gun.
    Then every r/c aircraft = drone.
    Then everyone claims AI systems when in fact they're just bog-standard (albeit-complex) if-then trees.

    Regarding the OP: just crunching data is NOT "machine learning".
    (RTFA and even they shorthand 'machine learning' into AI which it truly, truly isn't. Not even CLOSE.) This article fails to explain in any way how this experiment actually uses 'machine learning'. As far as I can tell, it's merely a broadly automated testing system where targeted variables are increased according to a pretty linear relationship of results. Essentially, it's just plain old Gregor Mendel's work, automated. THAT'S NOT MACHINE LEARNING, and it's absolutely nothing to do with AI.
    (I've tweeted to Will Knight asking him about both of these questions, and will reply here with whatever he responds.)

    --
    -Styopa
  14. Re:Give more Pesto to the People! by bugs2squash · · Score: 1

    if it's going to shine all day may as well make them 36 hour days, get 50% more pesto per day.

    --
    Nullius in verba
  15. Re:AGW is the opposite of ML? by Aighearach · · Score: 1

    Everything bad that happens is blamed on AGW, whether or not it makes any sense.

    Oh, great. Thanks Obama!

    (/s)