Funding Friday: KUMITA

#hacking education

Comments (Archived):

  1. JimHirshfield

    Hello Kitty next gen

  2. Mac


  3. Thor Snilsberg

    Looks like a great project. We need more stuff like this for all ages.To all the parents/grandparents/birthday present givers out there – the Kibo robot and Robot Turtles board game are great for the youngest programmers. http://kinderlabrobotics.co

  4. William Mougayar

    It looks good, as I’ve been looking for educational/tech toys like that. But the 0-3 age bracket seems a bit aggressive, no? Maybe 2-5 is more realistic?

    1. Twain Twain

      Realistic for N American households maybe.Chinese 3 year-old solves Rubik cube problem in 114 seconds:

      1. LE

        There are videos that show how to do that quickly. My stepdaughter (probably when she was 9 maybe) did it in less time I think. She just memorized something she saw on youtube. My point is not that it isn’t impressive for a 3 year old to do this but we don’t know if it was the first time she saw the cube or if she had been coached or had practice (do we?).

        1. Twain Twain

          “Practice makes perfect.”There are, though, some babies who grasp abstract concepts and skills faster than other babies. Call it intelligence, call it intuition. Whatever it is, each of our brains and skills are different from conception.In much the same way, some adults can think+say+do things faster and better than other adults.You and I couldn’t dribble a football past 6 defenders (at the same time) and send a spinning curve ball into the back of the net like Ronaldo or Messi. Neither of them have passed a Physics IQ test and yet they’re (practically) smarter than some Physics PhDs.Models of education for the very young are different between cultures. Not just N. Am relative to China and Japan, also with Russia.Note this: “(My parents) never queried why I was interested in maths and engineering – it was considered to be very natural.”*

          1. LE

            Neither of them have passed a Physics IQ test and yet they’re (practically) smarter than some Physics PhDs.This really depends on what the definition of ‘smart’ is. Or even better what the game is.For example if the ‘game’ in business is ‘make a great deal of money, be generally well respected, and not get into to much hot water’ (over the course of a long time, let’s say 40 years) then we can think of thousands if not hundreds of thousands of everyday business people ‘that are smarter than Harvard PhDs’ (the ones that teach at the bschool). My guess is that many of those people never even studied business in school either. They were ‘smart enough’ to figure it out on their own. I think this is a bit like the soccer player. They are smart enough to figure out w/o a full course what they need to know in a specific situation. (The guys who do towing of big truck rigs are similar. They are always referring to physics as part of how they move things. Years ago when I had to move an entire office all by myself I used physics I guess to move big things all by myself. Never took a physics course either…It was fun actually. I enjoyed figuring it all out and not getting hurt. Very satisfying.)A test is just that. A test. Some people don’t test well some people do test well but can’t easily apply that knowledge.

        1. Twain Twain

          I like Karpathy’s work. He’s one of the few AI researchers with a sense of fun.Nevertheless, his sentiment neuron is still at least 2000+ years behind my systems invention :*).

          1. Lawrence Brass

            That was what I thought!I know by experience that moving forward with R&D is slow, but do you have plans to showcase your work soon? I am getting anxious. 🙂

          2. Twain Twain

            Ah, I’d completed the R+D in expedited time because Twain brain had worked out EXACTLY the data sets and systems needed.As context, I recently got proof that as a banker I was 15+ years ahead of even Goldman Sachs in my innovation knowhow. https://uploads.disquscdn.c…In AI, I’ve had to be patient whilst cutting-edge researchers are only now waking up to the fact the data sets and frameworks are biased.This from Princeton University only last week:*…So now they’ll go at it for another 80 years, repeatedly using all the prob+stats tools (like their rating scales) that cause and/or amplify the biases in the first place.Of course, the only person in the world who’s already solved these problems is me, :*).

        2. Twain Twain

          That negative-positive scale of Karpathy’s is based on work from 2006 at Stanford that later got iterated on in Stanford Sentiment Treebank — which was the basis of Metamind, the startup founded by Richard Socher that got acquired by Salesforce and had raised $8 million from Khosla and Benioff when it was only 4 months old.Karpathy, Socher, Sutskever and Goodfellow are considered to be the stars of this generation of AI’s leading thinkers, by the way. They are all ex-Stanford and ex-Google (except for Socher).Check it out, I tweeted this to AI folks because they can’t even spot the Achilles’ heel in existing systems:https://uploads.disquscdn.c

        3. Twain Twain

          By the way, everyone should nominate me for MIT’s Disobedience award since I’m breaking at least 80 years of inadequate rating scales and 350+ years of probability’s strangehold on how we model human thinking, language and economic behaviors.And going against the conventions of all current AI approaches — lol.*…@wmougayar

      2. William Mougayar


        1. Twain Twain

          Yup and then little Chinese girls grow up to break the Rubik cubes (Matrix) of existing AI approaches to everything.And they wilfully defy non-sensical Professors of AI. Then one day AI’s leading research think-tanks go, “Uh-oh! Fake news! Toxic comments! Data biases! AI can’t understand our meanings!”Meanwhile, that female inventor LOLs because she already worked out how to solve it by breaking some maths+systems rules.And MIT launched a $250,000 Disobedience Award which she’d use to ship her system:

    2. ShanaC

      If you’re lucky, the earliest would be 6 months. A child needs to be able to sit up, crawl, and be able to have a basic hand-thumb grip to use this. Most kids can’t do that until around 10 months.I actually have concerns that a five year old may find this boring.…Five year olds tend to start thinking and behaving abstractly, asking lots of questions, including about math and logic/science. While the cards’ premise could theoretically spark growth in this area, the design setup looks and feels like something for younger children, since five year olds are also interacting with words, characters, and symbols.NB: people have told me in the past I should be a preschool teacher (or elementary school) especially since I’m curious about human behavior, and I tend to treat little kids with a combo of play and education. It pays super-poorly for something really critical to human development, so no. Still, I have thought about some of the problems involved with children and technology -and I actually wonder at the kindergarten/1st grade level, it might be better to teach coding concepts through physical body play and teamwork outdoors in a gym-like class with integration into other topics children that age need to cover socio-developmently, and testable topic-wise as well

      1. panterosa,

        Teach kids to think and problem solve consistent to their age level and they will understand coding because it’s logical. If coding were taught simply, almost like emergence at its root, then everyone would get it. But many people teaching it want to show you coding’s power, which means its complexity, and boom, there goes 80% of your audience out the door.

        1. ShanaC

          i’m not a huge fan of these sorts of toys. I generally agree with you. I also think that coding as an idea maybe should be delayed until around 5-6, maybe even 7, since I don’t think a 3 year old has quite the level of logic skills yet. (Piaget conservation skills are really needed if you want kids to understand the idea of a variable holding steps)

    3. panterosa,

      0-5 need free play. Full stop. They do figure stuff out on their own, if you trust them and let them run.

  5. Ivan B

    Brilliant idea – I pledged for one as well. It looks like a simple enough project that the chances of success are high, even if it inevitably takes longer than expected to ship to production.

  6. Mendelsohn

    Our 2.2 year old really likes Mattel’s Code-a-pillar which is quite similar to this, if perhaps a bit more accessible at his age. The build quality is questionable, but we’ve had good luck. Hard not to love all of these teaching toys.….

  7. karen_e

    Is someone suggesting that I need to learn to code beyond the ten HTML commands I used to use in WordPress?

  8. jason wright


  9. zakumanoff

    Looks neat. I backed a similar project called Kubo earlier this year. From what I can tell, Kubo takes a different approach to teaching “programming”. You have to program the commands (using tiles), and then you “run” it.

  10. LE

    This is interesting. Considering the team and the idea the goal is a bit unambitious. [1] $30,000 won’t do anything for them so when people (like this) do projects (like this) it’s more of a pr/marketing play in order to get attention. [2] And potentially greatly exceed the goal which will get even more attention. And it’s almost as if if they only raised $35k (exceed, say, by $5k) they probably would decide there isn’t a market for the product and idea.I think it’s cool and a neat toy and a worthwhile project. Seat of the pants impression, as far as what it does for kids in that age bracket, it’s hard to imagine that it will move the needle and lay the groundwork for what they state it will do. I am just not seeing anything but a really loose connection (given the physical game) to ‘teaches little kids the concept of coding/programming’. At that age with this concept. This could be tested with an older group prior to the ‘hello world’ to see to what degree it helps.[1] Don’t know that much about kickstarter psychology but is setting a low goal the equivalent of a .99 initial price on ebay? (I am guessing it is).[2] I think also projects like this might make it harder for others who might be ‘more needy’ to get what they need from kickstarter. In other words a slick professional team makes it much harder for the ‘little inventor or artist’ to succeed.

    1. Ryan Etheridge

      My first thought was pr/marketing as well. Is this really what Kickstarter is intended for – hoping for viral marketing via Kickstarter at little to no cost to the existing company? I’ve always thought of Kickstarter as a way for the little guy to get his toe in the door, find a market, build funding for a novel idea to compete against existing, award-winning companies.

  11. ShanaC

    That’s adorable and amazing, but I think they are stretching it when they say that this is appropriate for a newborn. At the earliest (and this might be a stretch too, infants don’t always have great attention spans either), you’re looking for a late 6 month old who is sitting up and crawling early. If your child can’t sit up and crawl, this isn’t for him/her. Ideally, your child should be able to grip stuff too, otherwise they won’t be able to hold the cards or the robo-doll.I also have some concerns about how an older infant/toddler will continue using this toy as s/he gets older. This isn’t a cheap toy, and at over $300 (list) for non-early bird buyers, even if I had a kid and was uber wealthy, I’d still pause. Children are hard on toys, and electronic toys with moving parts can (and often do) seem to break. Plus, without a vision/explanation of how this toy is going to be used as a kid grows up, or how it could be used by adults independently of their children,I’d wonder if this is going to be $300+ spent on something that is only going to be used for a year or two at the most. If I had a kid, I’d buy a much cheaper toy with longer lasting power if this question isn’t answered (duplos, for example, since their mechanism is designed to interlinked with Legos-and both I and my brother took advantage of this fact as children when we weren’t totally ready for the smallest of Lego pieces, but were also growing out of duplos).Still, I think the “making it a longer lasting toy” is a totally solvable problem. The really smart answer is figuring out how to create an open api to the doll, and allow people to create their own cards. If they are smart, they’ll host a user generated card marketplace, pay creators of good cards/card sets a fee every time they print and ship some cards from their marketplace , while controlling the cost structure, physical feel, printing process, and packaging of the cards. The marketplace also could have a “print your own without uploading to the public” choice too. This could encourage a broader audience, alternative uses, and ways for them to figure out how this toy is going to last for more than just toddler hood.Finally, although I briefly mentioned this: even with my other concerns, that $300+ price point outside of kickstarter early bird-land is crazy expensive. I really hope that they figure out how to get the BOM down (since I can’t explain where that price came from otherwise) as they try to sell this in larger non-kickstarter markets like amazon and toys-r-us, since sticker shock is going to be a HUGE problem for them.

  12. jschless

    price point concerns for many fantastic products similar to this. So many great ones out there.Market is growing at a nice rate.…Unit economics are impossible to drive down to accessible price point.Toys like these have drastic uneven distribution towards children in upper-mid class brackets.

  13. Alex Wolf

    Here’s a video trailer for the film on Kindergarten, and Mitch Resnick, among others, talks about the free play needed to develop creative thinking.

  14. Alex Wolf

    I’m about to launch a turd into the punchbowl. I hope you don’t mind Fred.This is exactly what kids don’t need, especially infants. It’s really for the parents – to make them feel like they are investing in their child’s future success. This will ensure nothing, and it won’t even do what Fred and others would be happy to support it for – as a coding precursor. It’s not the only slick electronic toy which is out there to make parents feel good. (And sadly make it look like people are investing in early childhood, which they are not, and they are wasting scads of money.)I work in designing for the early childhood space, directly on what children learn, how, and when, along side a psychiatrist cofounder. I can assure you that a child who cannot distinguish forms, cannot identify things, and certainly not know the word for it if they are pre-verbal are not grasping the big idea of coding. FORMS before THINGS before WORDS – is a key tenet the inventor of Kindergarten, Froebel, who sought to teach 3-7 year olds visual-spatial literacy before the learned to read and write. The upcoming Froebel documentary features Mitch Resnick. Froebel Kindergarten brought us Klee and the Bauhaus, Bucky Fuller and Frank Lloyd Wright – visionaries.It is most dreadful, and totally ineffective, to push the work of a 5 year old onto a 4 year old and so on down the line, in the effort to make the 5 year old look accomplished because he started doing 5 year old work at 4. This is what we have today.I am working on a totally different modality which aligns with all early childhood experts I’ve met, and many I haven’t, that actually teaches pattern recognition using the hard wiring in our brains from animal evolutionary survival skills.I’ve seen a lot of slick stuff at toy fairs and crowdfunding and all of it is hot air. Except maybe Sphero, but then you’d have to trust that a ball can actually teach a child a lot. And even then a ball doesn’t need to be driven by an iPhone to be a playful way to learn.

  15. ShanaC

    hahahaha.Basically, I agree with @panterosa:disqusCute, but not really appropriate. Stick with a flip book and make moo sounds at a picture of a cow with your kid when you read to hm/her. Both of you will get more benefit out of it than this.