When computers became popular, it became a new way to make media. You could make media before, but doing it with the help of a computer was 100x faster. There was immediate pushback - many creators did not recognize digital media as ‘real’. Those who embraced that media was made to be digitized (and created platforms to make that easier) pioneered new genres and business models. From those formulas of success, many winners emerged.

When the Internet became popular, it became a new way to share media. You could share media before, but doing it with the Internet was 100x faster. There was immediate pushback - many creators were very protective over their work being given away for free. Those who embraced that media was made to be shared (and those who created platforms to make that easier), pioneered new genres and business models. From those new formulas of success, many winners emerged.

So, what about AI? What verb will give new superpowers to creators? My guess is change.

AI is a new way to change media. There are hundreds of ways to remix something, but doing it with a model is 100x faster. There is pushback - creators are be rightfully concerned that their work is being edited & derivatives are being generated. Those who embrace that their creations were made to be changed (and those who created platforms to maker that easier) will invent new genres and business models. From those formulas of success, many winners will emerge.

But this new superpower is different than others. The ability to change the media you’re consuming near-instantly changes everything. It could mean everything we interact with (in the realm of software) can become hyper personalized on our prompts and behaviours. Every story becomes a universe, an entire genre. So this ability is exponentially more powerful than other waves. This is new territory, and it changes the relationship between people and their creative work, more so than ever before.

I try to spend a little bit of time understanding a tools affordances, where it can work, what it’s limits are, and then think about how it can be used practically in its current form, not in some future better state.

As an example - when Stable Diffusion was released, I create an app called Fresco, a simple game that to create artwork using diffusion models. But instead of trying to mimic hyper-realism (which, at the time, models weren’t good at), I made the app focused on creating abstract art, so that the hallucinations and inaccuracy became a feature more than a bug.

Sometimes I will choose to blend learning two things together into one project. For example, when I want to try a new framework or way of coding, I’ll force myself to use that framework in the creation of this new utility. In the Fresco example, it was an exercise in learning a new framework called Hotwire for interactive web app.

Creating an artifact of your learning that is more than a set of notes feels more satisfying too.

This post is an exploration of how this new capability is going to change storytelling communities. I focused on three native capabilities that were not possible before the breakthroughs in generative AI.

1. Creating stories on the fly
2. Talking with characters, not [just] fans
3. Building story worlds at scale

Let’s consider discovery and consumption. Finding media today is a process of discovery among millions of titles. This is made possible with machine learning on large amounts of user consumption data. AI now makes that discovery process almost unnecessary - you can craft a narrative on the fly. We can (and I predict will) go from finding stories from a finite (albeit large) catalog, to simply reading / watching the the stories you want. This changes the role of a consumer into a far more interactive role. This will also necessitate brand new economic structures for content.

But, this doesn’t spell the end of original content. Quite the opposite. I think most new content created will be of the derivative variety - offshoots of a canonical storyline, written by a human. I’m definitely biased, but I think this will be a golden age for fan fiction.

Fan interactions will be transformed too. Instead of merely discussing stories within fandoms, AI will let us also interact directly with the characters in a story. One could argue this isn’t new - roleplay is an established mechanism of fan fiction. What’s new is fans can now “talk to characters” from their favorite fandoms, in high fidelity, in depth, and on demand. This again, is a net add to existing experiences. Fans have an insatiable apetite for participation. People will still talk to other fans, but character conversations are a net new mode of interaction.

The most exciting part is the creative experience. There’s a lot of talk about AI taking away from artists’ control, but the reality is the opposite. Generative AI helps creators step off the infinite content treadmill, and elevates their the role. Instead of building a single storyline, creators are now world builders, inviting readers to play and create their own narratives. This is possible for every creator now, not just the JK Rowlings and GRRMs of the world. This mental model changes what it means to create “content” on the Internet.

It feels like we’re at the dawn of a new age of interactive stories. What an exciting time for creators and consumers.

• Eye tracking + pinch interaction sounds absolutely wild. Look at an app, pinch to select👌🏽. Very reminiscent of how iPhone touchscreen was 10x better than others.

• Feels like a new design space for accessibility. Can even imagine apps to help those with nervous system disorders (e.g. stroke victims) recover certain coordination functions, once more apps become a thing.

• Might not be intentional but enabling 2D apps in 3D is a great way to onboard existing devs.

• Naming it Vision Pro (vs. just Vision) is in line with a strong tradition of consumer vs. pro products. Also makes sense why Productivity is the leading message Vision Pro (my guess is “Vision” will focus on entertainment + “connection” features.

• Not hard to imagine an entire media ecosystem available “Only on Apple Vision”. Strong Apple TV+ play here too, possibly. Apple plays long games so this is probably many years out.

Lots more cool stuff here - https://developer.apple.com/videos/.

The world is a different place. COVID is relentless and the vaccines are a miracle. Feels like everyone’s lost someone to this virus. And yet, optimism is abundant. Especially if you’re building software.

The relationships we’re building at ‘work’ and ‘school’ are happening in software first, and in real life second (or not at all). It’s a good time to help humans to do more on the internet, and there’s so much to explore.

And because of this change, the day is regimented by our identity, not a daily commute to an office. Life just feels more fluid.

Flow is a state of complete immersion in whatever activity someone is doing; often signalled by losing complete track of time. You know when it happens, but it can be triggered. This article provides ten ways that contribute to flow state.

1. Eliminate all external distractions: put your phone away, turn off notifications.
2. Eliminate internal distractions. This is tougher but meditation and journalling help to clear the mind. Most imporantly, don’t try to enter flow state when stressed.
3. Work at your Biological Peak Time (BPT). Don’t fight it if you naturally get fatigued at a certain time of day.
4. Listen to the right kind of music. Repetitive, non-vocal music, that stretches for an extended period of time.
5. Work on one very specific task. Entering flow state with a list of things to do won’t work. Pick a task.
6. The task must be challenging enough, but not TOO challenging. If it’s too difficult, it will create stress and will be difficult to reach flow (see #2)
7. Have a clear outcome or goal. When you don’t know what you’re working towards, you don’t know when to finish, and procrastination loves it when this lack of clarity exists - it will push you towards quitting earlier.
8. Strategically consume caffeine. 200mg per day max (each cup is about 96mg). 400mg creates anxiety.
9. Stay hydrated. Not surprising but lack of hydration leads to a distracted state of mind.
10. Create a mental cue. A special sentence, action that you repeat every time you want to enter flow state.

Today, the structure of online courses is so much more, and some of the more successful ones I’ve seen have common elements:

• They all come with a supportive community of peers learning the same thing. Whether it’s a Slack group or Discourse forum, what you’re learning is just as important as who you’re learning with.
• They center around the instructor first, not the content. Twitter is an incredible learning platform because the act of Following is an indication of interest. Some people join courses today because the relationship with the course content starts with their connection with the instructor. Traditionally, you choose a course topic and then get introduced to the instructor.
• They encourage public accountability. The act of creating things in public (through writing and other creative fields) creates a natural skin-in-the-game.
• Built in status through exclusivity. Being a part of a (sometimes) restricted community creates a feeling of exclusivity. Completing the course and being a part of the alumni network carries social status too, just in the same way graduating from a university would.

In essence, moving to a default digital space in education means we’re seeing a lot of the shared support structures around courses that a “school” would provide be available to every course itself. It’s an interesting shift.

And while these platforms are really interesting, Twitter is the biggest passive learning platform in the world.

It creates a new learning modality not possible before - the consistent exposure to discussions amongst experts and those learning from experts. Anytime I’m interested in a topic, following Twitter users from that subcommunity or topic is a great way to immerse myself.

It also creates a natural graduation from passive to active learning through the forms of interaction - reading tweets, participating in conversations, creating new tweets about the topic. The exercise of doing everything in public creates a natural skin in the game for learning.

Although it’s primarily a social platform (can be performative, not always a safe space) - it’s ability to create a natural learning experiences is pretty interesting.

UPDATE: A tweet from Naval today reminded me of how lists can be used to shape Twitter into a learning platform vs. a social platform. You can create lists around experts in your areas of interest - and Twitter allows you (on mobile) to add them as additional tabs besides your primary feed. Really cool and underused capability.

Felt apt to start again by chronicling how it all feels. Since my last post, the virus spread is continuing. Many countries have managed to flatten the curve and places like Singapore have even stopped a second outbreak. But it continues is some of the larget countries. China is going back into lockdown. Brazil’s leadership is completely ignoring the virus and will quickly take over the US in cases. The US just crossed 100,000 deaths, more than the lives lost in Vietnam. India’s cases are rising, despite strict lockdowns and contact tracing infrastructure.

So, this isn’t over. Not even close.

We are back in Canada after a few months in New York, putting that life adventure on pause. Healthy, content, and planning for the future.

My daily routine has evolved. I’m up earlier, working out and meditating more. But the workday is intense. I’ve also noticed people are becoming very, very tired of this shit. The burden on everyone is huge and parents have a unique set of challenges. Most schools have cancelled the school year. Knowledge workers around the world are becoming accustomed to a digital by default workday. Childcare is becoming a real issue for families that are balancing it all.

Over and above anything else I’m thinking of the opportunities around us. Moving our spaces of learning and working into a digital space by default is probably the biggest societal change I might ever witness.

Sometimes this is obvious - they will tell you their pain, and it will be obvious. Most times it’s not and takes investigation - through conversation, and the learning from building products that don’t work quite as well.

Innovating from first principles is an exercise in serving a customer need, with the best tools available to you at the time. This cycle takes many years and many companies die in the effort. Some can do it once, but very few can do it again and again.