Krishna Rao

  • Just today

    This time of year we look back – and forward. We make new plans. Give ourselves the permission to renew broken promises. It feels good because the world has slowed down. You can breathe easy.

    And this is by no means bad. It’s great to have a vision, to imagine what you, your achievements, and life will look like this time next year.

    But once the vision is made, it’s time to shift your attention. From far out to right now. Not this month. Not this week. Right now.

    Here, you are taking actions. Here, you are facing your limitations. Here, you are experiencing defeat, triumphs, and, most of the time, just your plain old boring, precious life.

    The day’s not over (probably). Neither is the year. Breathe in. It’s time to work. Today.

  • My irrational belief

    Irrational beliefs can be used for terrible things – to start religious wars, vote dictators to power, or sell billions of the “edible food like substance” we call McDonald’s burgers.

    But they can also be used to do wonderful things. 

    Take my irrational belief. I believe that I will write a book one day, in the near future, soon.

    How do I know this? 
    I don’t. 
    Can I give you a logical argument why this is likely to happen?
    I can’t.

    For a few years now logic has held me back. How do you do a thing you have never done before?

    And then there are babies. Babies don’t operate as per adult logic. They take flights of beliefs – an irrational faith that their legs can hold them up when they can’t. And it is the very act of failure that builds the muscles that transforms them – from crawling monsters to erect maniacs.  

    And so, I have decided to be a maniac in this micro area of my life.

    It gives me a reason to sit in front of the laptop trying to flesh out a character’s motivation or the next scene when I really have no rational reason to.

    My irrational belief is that if I sit front of the laptop and stare at the blank screen for long enough, my mind will conjure up a semblance of a passable solution – and, on the rare occasion, dazzle me with brilliance.

    So, that’s my irrational belief for the next year. 

    What’s yours?

  • As simple as bread

    The aroma of freshly baked bread – or a delicate bite of one – is heaven on Earth.

    Bread can be made with just four ingredients – flour, water, a little yeast, and a pinch of salt. That’s it.

    I was reminded of this last night when I visited Marine Drive in South Mumbai. It’s a graceful arc of a road abutting the Arabian Sea. It was the night before New Year’s and the embankment was chockfull of people. Thousands sat with their backs to the road and faces to the sea.

    “I have never understood the point of it,” a friend noted. I nodded. I hate crowds and though I love the sea, I am not a big fan of Marine Drive. We walked for 10 minutes. A space opened up on the embankment as a large group got up and left. Our little group filled the space happily.

    We turned our backs to the road and our faces to the sea.

    A steady cool sea breeze fanned us. The traffic sounds faded. Our eyes took in the sea, breathing in the salty air. To our right, Marine Drive disappeared into the horizon. Despite the crowd, it felt like the whole wide sea had opened up just for me.

    I recognised the feeling. Just four ingredients. The Arabian Sea, a glittering road, our little group of close friends, and a pinch of salt.

    That’s all it takes.

    Heaven.

  • John Gruber disagrees with Mathew Inman

    And I disagree with John Gruber. Kind of.

    Mathew Inman the artist behind The Oatmeal recently wrote-drew an essay expressing his feelings on AI art.

    I want to start with a simple observation: When I consume art, it evokes a feeling.

    Good, bad, neutral – whatever.

    When I consume AI art, it also evokes a feeling,

    Good, bad, neutral – whatever.

    Until I find out that it’s AI art. Then I feel deflated, grossed out, and maybe a little bit bored.

    To which Gruber reacts:

    If your opinion about a work of art changes after you find out which tools were used to make it, or who the artist is or what they’ve done, you’re no longer judging the art. You’re making a choice not to form your opinion based on the work itself, but rather on something else.

    And then further:

    If an image, a song, a poem, or video evokes affection in your heart, and then that affection dissipates when you learn what tools were used to create it, that’s not a test of the work of art itself. To me it’s no different than losing affection for a movie only upon learning that special effects were created digitally, not practically. Or whether a movie was shot using digital cameras or on film. Or whether a novel was written using a computer or with pen and paper.

    I have two observations to make here.

    First, I think Gruber misinterprets what Inman means by the term “AI Art.”

    The way I see it, Inman is ranting against “art” generated purely by typing a few lines of prompt and choosing the prettiest option the AI throws up. When art is created using this process, then the AI is not being used as a tool at all!

    Second, I don’t think the art is separate from the artist or the process used to make it. Separating it, commoditizes art and is what most tech products seek to do.

    Suppose you walk into an art gallery and admire a piece of art on display. Would you value it more – emotionally, first, and then by how much you are willing to pay for it – if you knew that the artist took, say, six months to think through the work and made deliberate choices, large and small, to arrive at the final piece of work as opposed to writing five lines of prompt and put up whatever an AI tool churned out without modifying it even a little bit?

    Let’s take another example. Suppose you come across a beautifully designed pair of expensive shoes. Then you learn it is made by a sweatshop in Bangladesh – all legitimately, without breaking any local laws. Does that change how you feel about the shoes?

    I believe that the story behind what we create matters. That’s how we become fans of artists or why a Banksky is cherished far more than generic work of graffiti.

    But I agree with Gruber that we shouldn’t judge people based on the tools they use. But the key term here is “tools.” Inram too has no issues with people using AI as a tool of art.

    When you use AI as a tool, it’s akin to using Lightroom to process your digital photos as opposed to developing a photo film. It doesn’t reduce effort as much as it enhances it or allows you to work with entirely new mediums. And here too I agree with Gruber when he says, “Good art is being made with AI tools, though, and more — much more — is coming.”

  • When I asked AI to research ergonomic chairs

    It’s not often that you find yourself in the market looking for an ergonomic chair. And yet there I was.

    Last week, I spent half an hour a day, every other day on a little stool, as a physiotherapist kneaded my shoulders with an ultrasonic contraption to relieve a spasm. The cold metal didn’t seem to do anything – but it did not seem wise to contradict a medical professional. Plus, it was only mildly irritating.

    “Do you sit at your desk a lot,” the physiotherapist asked as though he didn’t already know the answer. “Yeah, I’m a writer,” I said squeamishly as gel hit skin. He didn’t say much during the session, but at the end of it, with all the seriousness he could muster, he encouraged me to improve my posture as I sat on the chair.

    He might have mentioned the words ‘support cushion,’ ‘quite cheap,’ and ‘Amazon.’ But I wasn’t really listening. My brain jumped straight to ‘ergonomic chair.’ But not a chair meant for muggles. If I was buying a chair, the investment would be quite hefty, and that meant – research.

    I skipped YouTube and Google, and like the cool kid I am, went straight to Gemini and asked, “What features should I look for in an ergonomic chair?”

    “That’s a great question!” gushed Gemini unlike any of my school teachers who always thought my questions were silly. “Investing in an ergonomic chair can make a huge difference in your comfort and long-term health, especially if you spend many hours sitting.”

    It then went on to describe some of the key features of ergonomic chairs, why they matter, and what to look for. Soon, I became well-versed with ergonomic concepts like lumbar support, arm-rests that moved in three dimensions, and mechanisms that adjusted seat depths and the space-time continuum.

    That evening as rested my butt on a variety of ergonomic chairs in stores, I felt like Einstein. The model I put my finger on came with dual lumbar support, a hefty bill, and a discount that was assured would expire by midnight (it did not).

    However, I wasn’t swayed. I hadn’t come across dual lumbar support in my research. And while I didn’t hurry back to Gemini, I did walk a tad faster for some last-minute research on this mysterious dual lumbar support.

    “That is a fantastic follow-up question,” Gemini said, “The benefit isn’t just in the “dual” (two) parts, but in how those parts provide dynamic and targeted support.” I was presented with a table comparing ‘Single Backrest’ with a ‘Dual Backrest.’

    This was all well and good, but something about the info didn’t quite add up. I decided to head to some of the sources Gemini had cited. The first website was Sihoo, “a leading professional manufacturer of ergonomic furniture.” The second website was of Hookay Chair, “also a leading office chair factory and ergonomic chair manufacturer.” The third website was Humanscale whose “award-winning office solutions, including self-adjusting seating, sit/stand desks, monitor arms and task lighting, inspire movement and support the user in every posture.”

    You see where I am going with this.

    Of all the websites Gemini sourced its information from only one – Canadian Centre for Occupational Health and Safety – could be deemed as having any form of independent academic expertise in ergonomic chairs.

    “Can you source this information from an expert or a research paper of some sort – instead of websites run by office products companies?” I instructed Gemini.

    Gemini deemed the request “fair” and proceeded to present me with information sourced from “ academic and institutional expert recommendations.”

    This time the sources turned out to be organizations that were “home of the globally acclaimed Ergohuman range of office chairs,” ones that offered, “fully customizable ergonomic chairs, stools, and workstation products,” and Herman Miller. To be fair, it also sourced some fairly inane information from the University of Pittsburgh, PubMed, and University of New Hampshire.

    Now, chances are that the information published on the website of companies selling office chairs is sourced from the engineers and health experts who work at said companies. But it is more likely to be written by copywriter who specializes in SEO.

    I am not alleging that the information presented by Gemini was inaccurate. Rather I want to bringing your attention to how taken I was by the authority with which Gemini presented the information to me. What felt like information grounded in solid research turned out to be information grounded in the blogposts of companies who are incentivized to sell products rather than disseminate research.

    As a former journalist and current documentary screenwriter, I am somewhat obsessed with knowing who or what my sources are. In the pre-AI era this wasn’t difficult to figure out since the information was often sourced from a book or a webpage. But now that we are firmly addicted to chatbots, it’s well worth asking them to cite their sources.

  • Why creators need better metrics

    One quirk – among many – of human nature is that we love to be loved. Not just by people we are close to, but the community at large.

    This must have helped ancient humans form and live in societies and obviously has its advantages.

    In the digital age, we seek this love, this validation, in Likes, Shares, and Subscribes.

    This presents a problem to the aspiring, and I’d argue even the established, creator.

    To aspiring creators it might seem like they are putting their soul on the line only to create paper boats and sending them to sail across a vast and rough ocean – with little to no response from digital denizens.

    On the other hand, established creators often find themselves saddled with this massive steamer held together by past successes that no longer reflects who they are today. But they feel like the only way to move ahead is to feed the engine more of what has worked for them in the past.

    Anchoring your measure of success (I promise that’s the last of nautical references) solely to reach and engagement metrics isn’t helpful for:

    1. Sustained creation when you are starting off.
    2. Experimentation and growth when you already have a sizeable audience.

    As the essayist Lawrence Yeo once observed: “If you base the quality of your work on its performance – and not how you felt while creating it – then you will never be confident in your own judgment.”

    Tell me if this sounds familiar. We set out to create something > It feels good to create it > The result doesn’t get enough of a reaction > The dejection supersedes the joy of creation > We stop creating

    But if we put Yeo’s advice to practice then we reflect on how the act of creation feels – and crucially give that feeling its due importance.

    Let’s call this an intrinsic metric and requires us to observe and value what feels good to us. This may not show up on any analytics dashboard, but as the author and humourist William Bruce Cameron puts it “Not everything that can be counted counts, and not everything that counts can be counted.”

  • Is AI a tool or an artist?

    Over at Substack Andy Williams shared this beautiful piece of stop-motion animation by the artist Victor Haegelin, for a short film called An Urban Allegory, directed by Alice Rohrwacher, and street artist JR.

    And this is what he had to say about it:
    “This is a stop-motion animation. It must have taken the creators hours and hours to meticulously and laboriously produce this animated street art. Why didn’t they just use AI? You know why, it would have been cheating. Show your workings, as your old school teacher told you.”

    While Andy is alluding to the ethical debate around the use of AI (one we certainly need to have), his post got me thinking about how AI is viewed in the creative workflow.

    In my experience, there are three broad aspects to most creative projects and the writers/artists/directors who work on them.

    1. Developing a taste
    2. Conceptualisation or imagination
    3. Execution

    Taste is the most under appreciated aspect of any creative product. It develops over several years and is the accumulation and slow composting of life experience and the art forms you expose yourself to.

    The conceptualisation of a creative project emerges from the marriage:
    (A desire to make some sort of impact) x (the artist’s tastes).

    And finally there’s the execution. That is putting the brush to canvas, hands on the guitar, fingers on the keyboard – you get the idea.

    Now, where does or can AI come into all of this?

    I think the use of AI in creative projects falls somewhere on a scale. On one end, AI is used purely as a tool, and on the other it is used as an artist – in other words to replace the artist.

    For instance, using AI, Photoshop can select subjects or remove objects in a photo on the click of a button. That’s using AI as a tool. But when you use Adobe Firefly or Midjourney to generate an illustration you are replacing the artist.

    So, what does it mean to replace an artist.

    1. You need to replicate the artist’s ability to execute. This is something that AI tools are able to do.
    2. You need to replicate the artist’s ability to conceptualise to create some kind of an impact.

    Here’s the thing. An AI tool has no taste because it has no limitations – it has been trained on pretty much everything under the sun and has no likes, dislikes, or quirks. After all one can argue that Quentin Tarantino’s sensibilities were shaped not just by the films he watched while working at a video rental store but also by all the films he missed watching because he was a limited human being.

    There is a reason why Leonardo Di Caprio or Matt Damon might choose to react in a scene in a specific way – this is shaped not just by what the director’s asking them to do, but how they choose to interpret it based on their unique, and hence limited, experience.

    All this hasn’t deterred “clients” from reducing the creative work they commission. But what I find is that the more they use AI to replace an artist, the weaker the emotional impact. Or rather the only emotions such creative work evokes is “How cool!” or “How pretty!”

    However, when you use AI as a tool – and this means a human with his/her limited experience shaping the final creative – you unlock all the efficiencies of a powerful machine without sacrificing the impact you want the creative work to have.

    I would like to close this rather long post with an interesting observation made by the fantasy author, Brandon Sanderson. Science Fiction and Fantasy as a genre takes the reader into fantastical settings that fill us with wonder – whether that’s Tatooine, Middle Earth, or the USS Enterprise. And yet, Sanderson observes that setting is the least important factor in what makes a fantasy story great. “A story that has a great setting but terrible characters is generally still a bad book,” he says, “But a story with a cliched and/or not that great setting with great characters still generally a fantastic book.”

  • Did someone order a virtual flavour maker?

    The road to innovation is not smooth. It’s certainly not straight. And sometimes it takes you down to nooks that are downright weird.

    Researchers at Ohio State University have come up with a device that can enable people to taste food remotely. Yes, I know, no one, absolutely no one asked for this stuff to be built. But innovators follow their hearts – or in this case their taste buds.

    In their abstract, the researchers note that VR and AR technologies are bringing the virtual and real worlds together. However, not enough progress is being made to replicate taste in the virtual world, they say with what I presume is a very real frown.

    Their solution: “a bio-integrated gustatory interface, “e-Taste,” to address the underrepresented chemical dimension in current VR/AR technologies.”

    The device comprises a sensor, which is inserted into food. The flavour information is captured and transmitted to a bunch of flavour packets that dangle from your mouth. Based on the flavour being captured, the packets release the relevant flavour mixes into your mouth via tiny tubes.

    The device releases hydrogels into the mouth based on the information passed on by the sensors. (Courtesy: Shulin Chen, via New Scientist)

    The device uses chemicals to simulate the five basic tastes that our tongue is able to perceive: sodium chloride (salty), citric acid (sour), glucose (sweet), magnesium chloride (bitter) and glutamate (umami).

    “Why, you might ask, would you want to taste someone else’s fish soup?” wondered the New York Times. One of the researchers, Yizhen Jia, responded by pointing out “that it’s commonplace to be able to see and hear what’s going on far away. Why not be able to taste it? Or maybe you’d like to taste recipes in a cookbook before you commit to making them. Maybe someday there might be a button on online grocery shopping services so you could virtually taste test different hot sauces before buying them.”

    Erm, cool?

    But to achieve this the researchers will have to find a way to replicate aromas virtually as well, since our olfactory senses (the sense of smell) play an important role in what we perceive as flavour. (Though not as vital a role as previously assumed.) And then there’s texture, which is also crucial to how we experience the food we eat. The researchers are aware they have a long way to go. “Everything has to come together for you to say, ‘This is good coffee,’” Jia told the New York Time, “A drop of chemicals on your tongue is going to feel different.”

    But the biggest improvement the researchers need to make is the interface. Dangling flavour packets from the mouth somehow makes me think of walruses – or characters out of Star Wars. Because when it comes to food, presentation is everything.

  • This man saved millions of babies – and you’ve probably not heard of him

    James Harrison, 88, recently passed away in a nursing home in Australia. He is estimated to have saved two million babies. All by donating his plasma every two weeks from the time he was 18 – all the way up to 81!

    Harrison is among a rare set of individuals whose blood contains an antibody called Anti-D. This antibody is important to save babies whose blood type is incompatible with the mother. As a result, the maternal immune system targets the blood of the foetus. This condition is known as Hemolytic Disease of the Foetus and Newborn or HDFN.

    The BBC reports that before Anti-D’s development in the 1960s, one in two babies diagnosed with HDFN died.

    It’s unclear how Harrison’s body started developing Anti-D, but some believe that it has something to do with the massive blood transfusion he received while undergoing a major chest surgery at age 14.

    Anti-D is administered to patients via injections that are produced using the plasma of donors like Harrison. However, research is ongoing to grow these anti-bodies in labs.

    (Source: BBC)

  • Writing in the AI age: More or Better

    Like many freelance writers, I am struggling to predict all the ways in which AI will affect my work – and how I can adapt to survive, even thrive, rather than become irrelevant. This can get rather overwhelming. But I believe wisdom lies in approaching such future-gazing like a jigsaw puzzle. While it will take a while for the full picture to emerge, parts of the new landscape will occasionally reveal themselves. Here’s one little nugget of insight. 

    I recently read a Substack Note by (presumably) a copywriter who managed to write a mind boggling number of blogposts, social media posts, newsletters, and other assorted creative assets in one month. I don’t remember the exact number, but the author claimed that the sheer volume of work would require a small agency to execute – and I was inclined to agree. Except this was produced and delivered by one writer who had developed a seemingly magical workflow that heavily deployed AI to deliver the work.

    Producing more is certainly one way for writers to survive the new reality. But what if instead of producing more, we were to focus on producing better. 

    For instance, let’s say, it took one writer 16 hours to produce one newsletter in the pre-AI world. Now, the writer can use AI tools to produce two newsletters that deliver similar levels of value and engagement. Or she could use AI tools to produce one newsletter in 16 hours – but with a higher degree of value and engagement. For instance, the writer could use ChatGPT’s Deep Research to weave in new data – and use Napkin.ai to produce a more engaging illustration.  

    Now, I can imagine that there will be clients who may not appreciate the better quality you produce. After all, the number of posts fits neatly into a spreadsheets in ways that quality does not. And so as writers we have a choice to make – to position ourselves as someone who uses AI to deliver more or better. 

    Maybe that’s the real jigsaw piece that clicked into place for me: This isn’t just about adapting a business model. It’s about deciding what kind of writer I want to be now that adaptation is no longer optional.