Mathscapes' Summer of 2019

> I think I understood the nuances of dealing with complexity.

I had a lot to say. And this is by far my most procrastinated task. And I would explain my journey of the things I learned in the words that I can try, along with my document of procrastination, knowing that I’ll laugh someday or judge myself in the future when I’m wiser. I’m glad you’re here and find this of interest.

May these words help my design process become. (better/grow)

A little about myself

I’m pursuing Human-Centered Design from Srishti Institute of Art, Design and Technology, 21 years old, a deductionist of decisions. I have liked Science, Economics, and Math in school but my grades let me believe I couldn’t pursue engineering (the working of all fascinating digital devices making decisions around me), that it would not make sense to me, and that there are wiser and smarter people out there more equipped for the job. I let that lead me to design, and I’m so glad about this journey. I switched two majors within Design to find steadiness and growth in Human-Centered Design. I believe that somewhere along the way, that fear of knowing not enough science and not retaining enough theory, to contribute to the field of technology has lingered and hindered my confidence a lot of times. But even so, the problems I find myself most curious about lie in the same domain. Fun!

Winter 2018, when it was time to apply for internships, I found myself unhappy with the traditional UX-UI work people were pursuing and calling for in the industry. The options available involving screen interfaces didn’t feel like something I wanted to engage with. I wanted to learn more theory, engage with the UX problem of rapidly-growing emerging technologies consistently shifting people’s lives and perspectives so invisibly. That’s where I wanted to apply UX.

On 15th April 2019, to my surprise, while I could only put the previous sentences together to explain what I wanted to work with, I was offered an internship with Mathscapes Research, and I am very grateful for the same. Thank you Capt Mashes! Before then, the decision to pursue research for Machine Learning felt all out of my reach, similar to school, having me feel that I probably cannot.

Calling this section: Poorly generalized conclusions that dishearten the real quest for knowledge, because hereafter, I observed where things weren’t working, calling for a lot of shifts in perspective.

Overview

It’s safe to say that I learned a lot about ‘learning’ during the course of my internship. I read an assortment of research papers, including A Mathematician’s Lament by Paul Lockhart, Tunnel of Samos by Tom M. Apostol and Usefulness of Useless Knowledge by Abraham Flexner, feeding into the necessity of exploration for holistic education, amongst many exciting leaps of knowledge in human history. I felt more curious about all the details and went blank-canvas to everything I came across, wanting to be open for all challenges coming my way. I was part of 3, almost simultaneous projects, a lot of reading, some making and the google machine learning course, alongside creating my blog website and using it to jot what I find joyous, for the next 8 weeks. Hello Complexity! I went on a quest to nurture new habits to regularity, including planning for each week and delegating myself tasks to meet weekly goals. I was a personal 7/10.

The first project, #AIForK12 (P1), now called Sigmoid, is my first formal opportunity to apply UX to emerging technologies. India is booming to become a tech-superpower, except this advancement is ignorant of how those ideas operate in schools. The brief of this project was to find the first, introductory element to Artificial Intelligence for children to engage with. Not all schools have access to digital devices, but all lives are impacted because of the power of computing. How might all school children prepare for the advent of artificial intelligence, given our current limitations to infrastructure and access? With a paper-toolkit, what if there was a way to still communicate the relevant ideas children could learn about early-on?

Parallely with Discoverdrome (P2), we imagined the blueprint of a school we foresee for the year 2101. It was curated, in every sense of the word, to the best of my knowledge and reflection to my own school experiences. Intended to be a week-long expedition, it had me play with material and imagine more and more for a 1cm:1ft scaled prototype as I built every new object- furniture, layout, possibilities. The need for modularity established an availability of choices in the space. There even were spaces imagined to be docked higher than the levels of tables and chairs. In hind-sight, those were elements compounding into ideas in the space, wouldn’t choices have children be intuitive? I was told that the intent of the project was to play with material and prepare for Sigmoid. But the project offered me a lot more. Speculating for the future demanded a reflection in the past, and to the growing importance and meaning of education over the years. Why else would a school need to exist so far in the future?

Education, my friend

My narrative to explaining this preface goes towards the human species. Man as a social being constructed meaning in his surroundings, that had to be universally approved and used correctly. It grew over the years, with discoveries and inventions, labelled, organized and practiced as bodies of knowledge and schools of thought. Complexity began to bloom! Before that, it was humans in an environment they knew little about, with the need to survive. Day and night patterns were observed. Threats in the surroundings were heard. Food from the surroundings were cooked and eaten. Materials were discovered. Tools were created, intended as one, applied at many. (Think of a knife.) Contexts began to inter-play. Newer words would form, condensing so much meaning and effort into a string of letters. But for any human to contribute, understanding the what, when, where, how and why would be the gist to grasp, often mistaken with knowing them instead.

Such a fascinating story, found overlooked in the space of schooling where our mediums to relate or find the relevance of education were limited, and our understanding was overlooked with mere presentation at the time of the test. And grades. If the prior may be prioritised, focussing on teaching me how I learn compared with how to learn, instead of every book they handed, things may have been easier. But even still, to explain to somebody what is worthy of their time and attention is an act of curation; benefiting their effort and time. But everyone in the space is different. What binds all humans, if not human virtues of play, creation and learning itself?

We learn to adapt with the nature we’re growing in, but we measure growth by our ability to compare, measuring trade-offs. The needs and desires we choose to solve and design for always have been because of the changing environment, and our need to keep track. We adapt new strategies, build new patterns that streamline the processes for us, often just to make redundant tasks effortless and less time consuming, playing out as different experiences for us, by us.

Education for Innovation!

Engaging with Complexity, fearlessly.

For me and the children learning AI. We’re humans and we look for patterns!

We’ve all looked up in the sky to see how beautiful it is. Each of us find it captivating in so many different ways, as analogies for so many dreamy thoughts. Was your journey still similar to mine?

1. I saw shiny points in the sky everywhere.
2. The more I look at it, I see more and more.
3. They’re not random! If you pick one and start comparing distances, you’d learn of patterns called constellations.
4. … Planets, and other celestial bodies also appear like stars… (or wherever your mind takes it)

Why the analogy?

Little observations like this came to inspire me for OpenHCI.

In my head, each star up there is holding the entire picture together. And despite it being a complex one, for some reason the aesthetic has not stopped us from exploring the sky for our own curiosities. We learn more as we engage for longer, try and wonder, purely by observation and play. What is formalised, is after a thought is defined, or when an experiment is concluded. It happens because you decide to engage, try to make sense and do what all humans do, find patterns! Patterns become the way of conveying meaning.

> The way, tareeka, approach, method or strategy of reaching a goal, meeting a definition would call for a plan at play, with different factors pitching their play. Some tareekas are more extensive, others time-consuming. The one more likely to sustain is the one that matters.

Sigmoid

1. How might we teach AI to school-children?
2. How might I understand AI?

Intelligence is a high-level word, employed in various contexts usually speaking of smart behavior. Digging deeper, it’s ability to work patterns within order and chaos is where it thrives. In order to buckle up for this problem, I think it was safe to assume I had to learn more to decide what could be the first thing to be introduced to children in AI. I’m still on this journey.

The word Artificial-Intelligence is easier on the word artificial (non-living things) to build association within children. Intelligence would have to be tackled in novel ways. I read theories on insect intelligence, looking at how primitive beings functioning on simple rules in nature inspire the processes that are programmed with algorithms. How ant colonies collectively act intelligent as they contribute to a shared, bigger goal, unaware of their purpose. Even Intelligence services, with their detectives building on clues observing cues use patterns to trace things backwards to the crime. There is inevitably a context, different guesses, attempts over leads, trials and errors that find them leads they stitch together to yield conclusions. They too build hypotheses and try to prove them with patterns that stand truth.

I was able to put across different concepts that emerge while engaging with learning about machine learning. But machine learning is programmed by humans, understood by humans relayed to the machines to execute. And the machine, a non-questioner, who merely recognizes magnitude and mathematics is the one that’s questioned. It felt like nothing made sense while everything did.

Takeaways

1. Simple rules act complex as they act over-time. They can be seen as elements, actors, tools behaving in its surrounding event after event, collectively emerging as intelligent behavior when an action is achieved at an expected time.
2. Computers are also trained with specific patterns, and ML now enables us to teach them how to find them when we don’t know them. The definition of intelligence keeps getting complex, employing different pattern building approaches yielding the same result. It still retains its nature so primal that it branches out to even define itself.
3. From start to finish, when a program is run, or when something is explored, the movement can be traced with a path. The route can be navigated by an algorithm, and the trajectory can be broken down and studied to understand factors affecting it. Usually defined by a user to a computer that is trained to ask questions. (I need you to understand I began comparing almost everything to my human mind, couldn’t help it)
4. ‘Making’ enables sense-making. And I think ‘sense-making’ utilizes if not all 5 senses, but more than just our eyes reading or watching, or just a thought registered in text stored inside of our head. I understood this with all paper prototypes explored that summer. When engaging with an object and the idea surrounding it, we utilise all our senses to put together the meaning we believe in, to understand it. I think that’s the difference between knowing and understanding.
5. If children understand how decisions are made in technologies they directly and indirectly engage with, they could empathize and offer meaningful contributions to their better presence in our surroundings. If life is about chasing more time to do what we desire, technology is a friend for the tools we have with us. But to sustain it’s positive presence in our lives, we should be able to intervene. How should children be made¹ to intervene?

(1. made: the past tense of ‘make’)

The Advent of Making

Make way for making!

As anyone who attempts to design, we recognize the responsibility it entails for the experiencing user. In design practice, there has always been a lot of emphasis on making, and doing. And those words alone make you wonder ‘how might I do something without being sure of it?’, followed by the seeming need to read more to be sure of it. Comforting at the moment to feel like you would be wiser next time, and undeniably you would be. But what if, a little extra effort to create something you’d want to remember more vividly might be more fun. And who said there’s no time in the world to read again later?

The perfectionist still wants to ‘procrastinate’ to deliver more effectively, waiting on the cue to arrive…

If you sit and consider cues to help you remember, you might think of signs. Signs in the surroundings: images and videos you watched, text you read, thoughts you had, connections you made, deductions, all stitching into another idea. So pure, and thoughtful, until your future self finds it obvious because time has passed. But was it really obvious?

> Make something of that information — a diligent reminder

If you managed to come across a pattern, congratulations! But does everyone understand it? Do they make the same sense as you do? How would you validate it?

> Signs can act as cues, fed into contexts, to instill play and curiosity. What if what you make, in the longer run, acts like a cue to your next expedition?

Some days, I was thrilled with this idea of engaging with a new material or mode to express ideas, I would jot random thoughts on post-its around me as they would come. In the moment, they were just sentences, later growing into ideas.

On other days, I saw the need to be managed. (Dramatically, I was in full belief that I needed to hire a manager. How ignorant, I needed to understand myself better.) When all things felt spilling over, unable to see what bore out of my attempts as other tasks were left behind. The weekends felt blessed, and fortunate to make use of more time and unwind. But if someone else would just do the thinking for me, I might make better at the same time. Better yet, if someone would do the making, I may be able to think better. They were seen as separate tasks, trading off my time and energy. Because when you’re invested in one thing for too long, you tend to forget, and simple things start to grow in size.

> If you sit and consider how much work making is and feels like, with the need to plan and organize everything else to execute it, you’d wait for a little more motivation, a little more self-assurance.

> What if instead of comparing small leaps of growth, the focus could shift to comparing greater milestones?

This calls for a mere shift in perspective to do the same work. It’s a different looking way, method, approach, strategy, whatever you want to call it. But if you break it to compare with my previous approach, I think this covers more ground if everything is given enough time. In complete retrospect, as I look back at my process, I realize the extra reading time I would demand and employ did bring me progress, but the effort put into creating artefacts (mind-maps, post-its, plans, sketches, timetables, and craft prototypes) were the bigger wins. Not so much the knowledge I think I hone, but the work I did (which does include the knowledge) that I can look back at and find worthy of the time I spent on it.

Which brought me to look at time as a limited resource, stripped off its usual y-axis, asking to be managed. And we say hello to mathematics once more.

> In business and other places, they say in order to manage something, you have to measure it.

If you measure consciously, you may be able to stick with your plan of action. But Actionable is also a verb, anything more than just a thought in your head. Logically dictating me back to making/creating with what I know. (I hated the word make for how it would always come a full circle) I observed how my plans were written. How much the language and the words I used, the detail in the tasks I’d pile, and the time I seemed to give but consistently felt like I couldn’t do enough. Words became critical. Explanations, necessary. And observations — documented, a must. My blog saw more writing than I ever did in the past 3 years. I questioned the word ‘code’ itself, how naive it sounds placed in ‘color-code’ or ‘code-names’. We’ve all known that words often can be used interchangeably, as long as their place in the narrative is justified. I wondered if concepts in Computer Science have either been made to look complex, or the lack of the right progression has us choose to address them on a more surface-level. That complexity and the fear of wrong, instead of questioning the patterns behind the literature we’re part of is what makes us uncomfortable. And it comes down to focus on using the right labels to the words, instead of getting to them, tracing their paths, structures, and behaviors organically.

The unseen fallout

For a definite period, I felt out-of-tune with the words I was using to communicate. My points wouldn’t go across, I would sound too high-level, and it made me question this journey once more. Our ever-evolving present is so dense in information that the simplest way of representing them is language, who I formed a love-hate relationship with. Amongst all constructs of the began criticizing language because all things I studied were high-level concepts being translated into low level, as functions one after the other. Meta-processes, and processes within processes. It was comforting to imagine from coding that they can be nested functions, but it was frustrating to think that my understanding was faulty because people couldn’t make sense of it.

But I was only trying to make sense of everything in relatable ways. I could be oversimplifying complex information and misinterpreting the complexity of what I was studying, but I more strongly believe in the fact that we have too many synonyms in different contexts that basically mean something you might already be aware of. I still think children would be more open to learning with simpler words, because beyond a point it comes to questioning whether it’s really something new or the reiteration of the same information? And when it comes to that, we would all want to validate our understanding with a different sense (making would help), if not text that I have been critical about. Because the use of just words is tricky, they associate and build on top of concepts, letting us overlook the systems at the hands of a word, text or spoken, bearing the load of all that went behind it.

Patterns to Play with Time — The Relevance of Algorithms

Get comfortable playing with measurements | Comparisons inspire Optimization | Free space off redundancy! | I did not know this was coming.

> I have gone from ‘being fascinated with the parallels in machines and the human mind’ to recognising that both are just performing math, by running on algorithms.

To understand machine learning, it’s the mathematics that needs to be made relevant and approachable. And shortcomings of schools aside, our interactions and decisions in our surroundings are also calculated with time. I think time, not only fundamental, is also our first metaphor for numbers, offering a check on every element at play within a chosen context, limiting their participation and affecting their behaviors in the space. And it runs on counting, so there’s the math. I like calling this idea time-transactions. People understand their losses or gains over time.

The moment we keep limitations in capacity and time, there are economics at play that demand pre-planning for both algorithms and humans 1) to find and employ ways of prioritizing and organizing information in a pattern 2) with choices that ensure efficient use of resources and 3) minimizing redundancy. Understanding of things is shaped better when assessed against an axis of time.

Planning then began pointing to optimizing to buy time, if not deliver within a time constraint. It pointed to rearrangement and grouping of information, to organize factors that affect so they are 1) understood better while setting them and 2) being made easily accessible later, for meaning that can be called timeless.

Endnote

I wrote all that I think I needed to say because in my head it’s all chaotic and this is unburdening. But apart from that, I think the reminder of planning and making has always found itself in tune with the nature of algorithms, and this was an attempt to document and acknowledge the same.

Acknowledgments

Biggest thank you to Gaurav Singh (who will review this and help me understand where I’m at with all these words stringed in this manner in my head), my mentor, for initiating this quest that I continue to find intriguing each day. I’m only 22 and always running from things that don’t make sense, this has been a journey of a lot of Mathematical Play that I would otherwise be too afraid to be trying.

My parents, who were the OGs of planning and upbringing me, as I would be kicked out of the house and only be given 2 hours to study and spend the rest of my time learning new things. I find myself liking the latter more but the former with constraints on timing and regularity were the things that had me thrive as a child. I’m working on lifestyle changes for a good sleep cycle. (Did you know, short-term to long-term memory happens over sleep?) I might not be able to stomach this project if I don’t sleep well :o

Vineeta Rath, for reminding me of all the BodyFirst mantras. They’re algorithmic and I see it.

My friends for all their support. They helped me get back up each time I saw no hope. I hope I do all of this justice.

About the author

Simran Singh is a human-centered design student pursuing her undergraduate degree at Srishti, Bangalore at the time of writing this. She identifies herself as an analytical-dreamer, her work focuses on understanding the use-cases of emerging technologies and applying it in a user-friendly context. She sees value in algorithmic thinking to help understand, decode and apply at the design context