Podcast To Listen To: Ayah Bdeir on Positively Gotham Gal
Ayah Bdeir is one of the most impressive entrepreneurs I have come across in the last ten years. The Gotham Gal invested in the Little Bits seed round in 2011 and we have watched her build that company over the past six years. In this podcast Joanne and Ayah talk about the early days of Little Bits and the power of people believing in the vision early on.
Listened yesterday.Big fan of this founder.Product of a father that made us as kids head out on junk day and strip radios and toasters and soder things together that did stuff.Still think that the positioning of Little Bits should be broader as their market honestly should be each and every parent at some stage of their kids growth.
THIS: “When I started the company, I was like ‘We’re late! The market’s already fully-developed! But the reality was that it took five years for the market to catch up to the idea…”Everyone imagines that Google, FB, IBM Watson, Apple, MS et al already own all of AI and that some AI verticals are also saturated by startups, e.g. image recognition, speech recognition and calendar assistant apps.The reality is that the market is only now waking up to 4 major problems in AI that the big techcos and startups, alike, haven’t been able to solve and can’t solve(*):(1.) Natural Language Understanding => AI can’t support decision-making or filter “fake news”;(2.) Data and algorithm biases => affects search results and product recommendations;(3.) Lack of explicability by “black box” Deep Learning systems => financial risks; and(4.) Limitations of classical computing => with Quantum Computing on a 2-5 year horizon and IBM already released Q Experience system.—–(*) Can’t solve because it would involve the re-engineering of about 350+ years of Scientific Method and 2000+ years of Logic.(*) Can’t solve because language is subjective and, as Ray Kurzweil pointed out in 2006: “That (subjectivity) is not a scientific concept.”So it would take an “inventor of inventors” (or a crazy person) to take on 2000+ years of maths and science to invent a system capable of making subjectivity as scientifically measurable as objectivity, and to create the basis for Natural Language Understanding; human+machine coherency and the ability in economics to model irrational human factors alongside rational human behaviors and events.
Another interesting point Ayah shared was that where we study can influence what we end up doing. In her case, MIT MediaLab’s MO of taking apart things that already exist and then putting it back together.In my case, my alma mater is where Rosalind Franklin did her seminal work on DNA, Maxwell discovered electromagnetism, Peter Higgs did his research on the “God particle” and my first year maths lecturer wrote ‘Supermanifolds’ on multi-dimensional space.All of these things inform my systems approach to solving the data signal:noise problem in AI, especially Natural Language Understanding.In June 2017, two papers were published that made me very happy:(1.) Qudits for quantum computinghttp://spectrum.ieee.org/te… https://uploads.disquscdn.c…(2.) The brain is multi-dimensional and involves super manifolds — this research fundamentally updates the simplistic neuron from Rosenblatt’s 1957 work that is the basis of today’s Neural Networks.* https://eurekalert.org/pub_…
The payoff of products like lilbits / grade school coding will not be in computer science itself but rather it will be in upcoming productivity explosion of engineers and scientists.
Liked it!Ayah is a role model young woman entrepreneur.
Interesting.For the business, if can sell digital kits, fine.=== EducationBut it appears that the issue of more concern is education.Just what the heck to do with/in education is challenging. Or, education is about preparing kids for the future, and, right, drawing from Star Wars, “Always difficult to see the future.”The conversation had a lot of emphasis on curiosity: Yes, commonly kids are very curious, and the best explanation I heard was that the kids saw that the adults had all the power; the kids felt subordinate and dependent and didn’t like it; so the kids were highly motivated to get the power of adults; for this, the kids guessed that they needed to learn more so had curiosity.Sure, there is a flaw for the explanation and maybe for the kids: Not nearly all the power of adults is from adults being curious, at least not curious about how little electrical, electronic, or mechanical things work!Moreover, a lot of curiosity has a downside: (A) Some curiosity can be about subjects that really seem to be nearly irrelevant to anything else in life. (B) A kid that is really curious can neglect other topics, subjects, etc. that are known to be important.But, sure, a curious kid is much to be preferred to a bored one!One approach to learning for kids is to emphasize the 3Rs — reading, writing, and arithmetic. Okay.=== TechnologyFor technology, what to emphasize? I’m uncomfortable saying that we should push kids very hard on technology because it tends to go out of date — at least become much less interesting and valuable — so fast. E.g., at times in the past, new, advanced, powerful, valuable technology was working with animal skins to make clothes, then fiber to make thread and cloth, stone to make spear and arrow heads, wood to make buildings, tools, and wheels, metals for tools and weapons, stone for buildings, wood for ships, iron and gunpowder for guns, steel, steam, chemical engineering, right, as in the famous one word of advice in the movie, plastics, electric power, wireless communications, radio, telephones, TV, radar, transistors, nuclear power, digital computers, computer software, digital communications (e.g., for the Internet).One reason this list is so long is that such technology quickly has the profit margins fall to very thin. The technology, e.g., working with animal skins, working in wood and stone, are still around, but we are not thrilled to see a young person seek to tan hides or be a carpenter or stone mason.Yes, I know: Even tanning can have some value! E.g., supposedly on the recent trip to Europe, at least once our US First Lady Melania carried a handbag from a company that has some handbags from some animal skins and that sell for $100,000+. So, there can still be some money in animal skins!Generally it appears that technology gets pushed hardest by (1) national security and (2) medicine.=== What to EmphasizeSuppose some kids want to work with technology in part with the idea that it can be a good foundation for a good career for life. So, what should the kids emphasize? Tough to know. Here is my guess: In one word, fundamentals, and that because they (A) will always be essential and (B) are the single best way to know what is impossible, what might be possible, and for the possible the keys to making it real.Sure, the important fundamentals are not always old, e.g., DNA, CRISPR, etc. in bio-medical technology.I’d be reluctant to see a kid get all wound up about, say, carpentry, say, making little boxes and toys out of wood, glue, and paint. Then, sorry ’bout that, I’d be reluctant to see a kid get too wound up about low level digital electronics.Why? Because now IIRC, digital electronics can sell processor chips, each with, say, 100 million transistors, quantity one, retail, for about $10. The last time I went computer shopping, there was an 8 core AMD processor with 64 bit addressing, error correcting memory (ECC), and a 4.0 GHz clock for less than $20 per core.Net, that stuff is cheap, dirt cheap, and rapidly getting cheaper.Gee, the $20 is not even a dinner at McDonald’s for two teenagers in love!But, that AMD processor can be enormously valuable if can think of something good to do with it.E.g., that processor is capable of being one heck of a Web server, sending maybe 50 Web pages a second, with several ads per page, for maybe $100,000+ a month in ad revenue, assuming, of course, that the Web server is providing something good enough to get that many users.Astounding.So, then, from this example and more, we begin to suspect that what matters is not low level digital electronics, say, with AND, OR, and NOR gates, adders, instruction decoders, micro-code, virtual memory, page-segment tables, translate look-aside buffers, n-way set associative cache memory, cache invalidate, etc. but what is done with microprocessors, etc.For that, the next step is software. So, in choosing what to emphasize, between the hardware, e.g., that AMD processor, and software, emphasize software.Next, software from individuals and small groups is well on the way to being boring, very routine, writing code as mortar to bind together API calls where the real work usually is and will be.Then, there is a challenge: How to organize and manage successful software projects of dozens or hundreds of people. So, for a cliche, move up to systems level thinking: There can be important work to do here in system design, reliability, testing, documentation, revision, etc.Next, I’m reluctant to emphasize software too strongly because what’s known now is quite routine, and, for more value from microprocessors and software, the challenge is not how to write the code but what code to write.The first need here is, say, to identify a customer want/need, e.g., mobile computing with, say, smartphones. But I’m reluctant to emphasize this: Big wants/needs are easy enough to identify; the problem is that for the unmet wants/needs people with just software don’t know what software to write, that is, don’t know how to write the software to meet the customer want/need.Then, here is a guess of mine: Software takes in data, manipulates it, and reports results. The main question for the unmet wants/needs is what manipulations to do. Well, all such manipulations are necessarily mathematically something, and, then, by far the best path to discovering the needed manipulations is to proceed mathematically. Examples include GPS, Internet routings, error correcting codes, and cryptography.So, instead of low level digital electronics, microprocessors such as that amazing AMD thing, and software, I’d recommend trying to move up to systems level considerations and, in particular, mathematics for how to do the needed crucial, core data manipulations.At one point, I was surprised to discover some of what the academic electronic engineering departments were doing: They had been teaching students how to make use of resistors, capacitors, inductors, vacuum tubes, and transistors. Then they moved to careful consideration of Maxwell’s equations.Then apparently they concluded that that material was too old and too short on potential for powerful, valuable things for the future and moved to much more in mathematics. So, their view seemed to be that the electronics they had been teaching needed new work to do, new tools for new problems to solve, and got good at some nicely advanced applied math — stochastic processes, stochastic integration, stochastic differential equations, classic and non-linear filtering, and deterministic and stochastic optimal control. That math needs to be done carefully with fully detailed theorems and proofs, and surprisingly, for the good texts and papers, the electronic engineering departments were often the leaders and way ahead of the pure math departments. Amazing.=== CreativityWe can believe that the valuable future will be from things that are new, powerful, and valuable. Well, then, usually these things will need some creativity.And part of the discussion was creativity. Well, if we want new things, then we are essentially forced to look for some creativity. And, yes, it’s easy to believe that curiosity is a good aid to getting to creativity.One of the problems in education is manipulating the students to work for praise from teachers and parents. In this way, some students can make excellent grades in material actually they don’t much care about. The result can be a student short on both curiosity and creativity. Not good.Education has problems fostering creativity. Buried in his The TeXBook, D. Knuth has:The traditional way is to put off all creative aspects until the last part of graduate school. For seventeen or more years, a student is taught examsmanship, then suddenly after passing enough exams in graduate school he’s told to do something original. In practice, there can be a serious conflict between the students who do the best on the exams and the students who are the most creative in the sense Knuth is talking about — publishable academic research. That is, having students diligently devoted to straight As in K-12 and college and also good at research in a Ph.D. program can be difficult; sure, each is difficult, but getting both in one person is more difficult than it should be because to some extent the two goals conflict. But, academic research aside, Knuth’s concerns can hold for creativity more generally.For me, Dad had a Master’s in education and, there, picked up a lot of the far out thinking about education. As a result, for me he emphasized unguided curiosity and learning and never emphasized grades. As long as I learned and passed, he didn’t care what the grades were. The “unguided” part was not so good.Then in graduate school, what totally saved my tail feathers was some creativity: At one point I took a problem I’d seen in a course, confirmed that likely no solution was known, got a reading course to study, not necessarily solve, the problem, and dug in. It was fun: In the evenings my wife was watching TV, and sitting next to her I worked on the problem. I had some ideas, found a solution, wrote up the work, and turned it in, all in two weeks. Fast course!Some of what I proved was surprising, and word of my success spread in the department.The work was obviously publishable, and later I did publish it.The work was good enough that any of the faculty members would have liked to have done and published the work.So, what I did was creative in the sense of Knuth, etc.Soon I discovered that in the department I had a polished halo. That was good to have.So, if go far enough in academics, then creativity can be nicely rewarded. But to get that far in academics, might need to slog through a lot of really boring stuff!So, yes, emphasizing curiosity and creativity can be good and even crucial if get to a Ph.D. program and for some cases of entrepreneurship, but in much of K-12 and college curiosity and creativity can conflict with the system!=== What to Teach KidsFor kids, at times I’ve wondered what to teach them. Well, not nearly just math! Here is a list: Emphasize their development, e.g.,academic, artistic, athletic, creative, emotional, empathetic, entrepreneurial, ethical, mechanical, moral, psychological, quantitative, rational, religious, romantic, scientific, social, technical, verbal.There is the oldJohn Dewey, Democracy and Education: An Introduction to the Philosophy of Education, The Free Press, New York. IIRC, somewhere in there Dewey defines education as the passing down from one generation to the next.Well, what the next generation should do is create some good new stuff, and that stuff itself, since new, can’t be so passed down. So, really, instead, what need to pass down is something about creativity. The usual K-12 and college are not very good at that!
In my opinion Ayah Bdeir is a living legend. We all know his struggle and journey to reach this level. I want to pick up his journey in my essay writing. She becomes an idol in a young society.