🌱 5-Bit Fridays: The AI unbundling, startup lessons from Taylor Swift, how to work with designers, adding value early in your career, and the petition to halt AI development
Edition #20
👋 Welcome to this week’s edition of 5-Bit Fridays. Your weekly roundup of 5 snackable—and actionable—takeaways from the best-in-tech, bringing you concrete advice on how to build and grow a product. Catch up on our other editions here, or our main deep dives here.
Happy Friday, friends 🍻
Exciting week content-wise here at How They Grow! On Tuesday, I posted the first edition of our new spin-off series, Why They Died. In our first trip to the startup graveyard, for a post-mortem on what led to failure (and the lessons to be learned), we took a look at…Vine. If you missed it, we break down four reasons behind the fall of TikTok’s predecessor.👇
Otherwise, let’s get into it.
Here’s what we’ve got this week:
The Idea Propagation Value Chain, and how AI is unbundling it
How to work with designers: Concrete tips on building rapport and mutual respect
If you see a problem, solve it: Adding value in your early career.
Favor: If you enjoy today’s post, please click the little heart button to like it—it helps more people discover my writing—or share How They Grow with a friend. 🙏
The Idea Propagation Value Chain, and how AI is unbundling it
In a recent essay of his, Ben Thompson of Stratechery introduces this concept of The Idea Propagation Value Chain.
Simply, it’s how ideas spread, and it goes like this:
Creation: You imagine a concept for a new video game
Substantiation: You have to articulate what you’re thinking, by putting the idea into a format for other people to get. Could be verbal, visual, written, or whatever other medium gets it into a digestible format.
Duplication: Whatever format your new video game idea is now in, it must be replicated. A hand-written note, for instance, needs to be photocopied or pulled into a shareable word doc.
Distribution: Your substantiated idea now needs to be spread far and wide. The internet makes this wonderfully easy.
Consumption: People discover your idea, process it, and now can imagine your new video game, get excited, and share it.
In other words, our productivity in the world is really a function of this chain. Communication is everything.
And since the days of hunting wooly mammoths, Ben makes clear that the history of communication has always been about removing the "bottlenecks" in this chain. At first, ideas for new hunting techniques could only be shared orally around campfires, which obviously capped the number of people who could hear them. Consumption was the bottleneck.
Then cave paintings, stone tablets, and writing came along. Each bringing improvements in the substantiation of ideas, as well as unbundling consumption — increasing the number of people who could consume an idea. Now duplication was the challenge, as it had to be done by hand. The revolutionary printing press then solved this issue, unlocking huge duplication/distribution improvements. Ideas could now spread far beyond the source.
But, distribution still required time and money — which made it a bottleneck for profitability. The internet removed it by making distribution free and available to everyone.
So, that leaves us with the only remaining bottleneck — the creation and articulation of ideas. But, as Ben argues, this is also on the verge of being removed.
AI is amplifying our creativity and articulation
AI has—to use an overused term—democratized creation → substantiation. And it’s obliterating this top-of-chain bottleneck at an alarmingly fast rate.
Going back to your hypothetical video game idea. Even just a couple of months ago, where would you have started? 🤔 If you can’t write code, certainly not with a proof-of-concept you whipped together. Maybe you’d try your hand at drawing something out or creating some sort of concept board. Most likely, you’d write it out.
The point; getting your idea articulated to a place outside of your head so it can be consumed, non-verbally, faced a ton of challenges if you lacked certain hard skills.
Now we have AI tools coming out the wazoo. Many are breaking down that barrier.
Generative AI can take your loose idea (or come up with one for you) and give you a detailed spec for how the game should work. It can even give you a plan for creating and marketing it.
Beyond that, you can get the code for an MVP, and even build an entire game, all without any programming background. And the same thing applies if you have an idea for an app, Chrome extension, or even a basic website.
Then there are the image generators like Midjourney and DALL-E. Instead of leaving a concept up for interpretation, you can show what you’re thinking with just a few prompt iterations. Imagine you’re a car designer, and have an idea for a futuristic electric Land Rover, with solar power from the roof. Well, in less than a minute, your idea exists. And with options.
Or, take these creations by Nick St. Pierre from Midjourney.
Imagine having an idea for a show, and being able to create an entire storyboard off your script. White Lotus vibes.
Or, being an interior decorator or wedding planner, and quickly being able to whip up concepts to show your clients. (Or, doing it yourself of course)
This leap in productivity we get from severing the connection between creation→substantiation is truly insane.
And it’s easy to see the impact of this across fields. Take an example of a non-technical founder with an idea:
They can now go build prototypes of their idea without needing to beg, and/or convince an engineer to help.
Meaning, you can create ugly working code that you can use to attract a technical co-founder. (An aside: I was chatting to a friend who’s an engineer recently, and he said the best way to get a dev to help work on your new idea with you is to bring them some bad code and ask for their feedback. They can’t resist helping fix it. 💡)
But, we're not far off from a day where AI simply codes an entire product for folks who don't know how to code, which blows the field wide open for anyone who wants to build a technical product.
Now, changing gears from generative AI to Human-generation.
Specifically, to Taylor Swift, and what she can teach us about startups.
5 lessons on startups from Taylor Swift
You probably don’t know this, but I’m a huge Swifty.
So, when I saw
's post, "What Taylor Swift Can Teach Us About Business”, best believe I read it immediately. 😍One of my favorite things is when people distill lessons and insights from places that others wouldn’t typically go looking for them. And zooming in on Taylor Swift for startup advice is definitely unexpected.
In Rex’s post, he brings us 5 business lessons we can learn from Taylor’s impressive career, namely:
Great Products Expand the Market
It’s All About Community
Brand Building: Is Brand a Moat?
Product Velocity Wins
Reward Your Superfans
In today’s post, let’s just take a look at #1.
Great Products Expand the Market
I’ll let Rex share his insight:
Taylor Swift was born in Pennsylvania, but at age 13 convinced her parents to move the family to Nashville. Determined to be a country singer, Swift went door-to-door in Nashville to the major country music labels—which all rejected her. Undeterred, 13-year-old Swift began to write her own music and perform that music in bars around Nashville, hoping to be discovered.
The mechanics of the music industry are fairly similar to those of the venture capital industry. Music labels have A&R reps (A&R stands for Artists and Repertoire) whose job is talent scouting—they’re out and about looking for the next big thing. Once a label signs an artist, it offers an advance—essentially money to fund the production of an album—and the label then contractually owns the rights to recoup that advance and eventually profit. The industry is much more complicated than that, but that’s the gist. Like venture, music operates by a power law in which most artists fail, but the ones who make it (like Swift) more than make up for the ones who don’t.
In Nashville, RCA Records discovered a young Taylor Swift. But there was one problem: they didn’t want her to be a country singer-songwriter. RCA looked at the data, and they saw country music on the decline with young listeners; they looked at Swift, and they saw a teenager they thought they could mold for a more lucrative segment of the market. They could give Swift other people’s songs to sing, and she could be a vessel for the label.
They were wrong, of course.
Swift rejected RCA’s offer, unheard of for an up-and-coming artist. Swift later reflected, “I didn’t just want to be another girl singer. I wanted there to be something that set me apart and I knew that had to be my writing. Basically, there are two types of people. People who see me as an artist and judge me by my music. The other people judge me by a number, my age, which means nothing. It’s not really a popular thing to do in Nashville to walk away from a major record deal, but that’s what I did.”
Swift later signed with Big Machine Records, and the rest is history. (Swift’s catalog constitutes about 80% of Big Machine’s revenue.)
A country music-era Taylor Swift
The key insight here is that RCA underestimated the market. Yes, country listenership was declining among young people. But great products revitalize stale markets and ultimately expand their markets. Swift did both, first proving that there was an opportunity in country music, and later growing beyond country to become a bona fide pop star.
Swift’s story reminds me of a fatal error in startups: underestimating market size.
A classic example is that of Uber. As Bill Gurley outlines in a great post from 2014, many experts dramatically underestimated Uber’s market opportunity. One finance professor at NYU, Aswath Damodaran, wrote: “For my base case valuation, I’m going to assume that the primary market Uber is targeting is the global taxi and car-service market.” He arrived at a TAM (Total Addressable Market) of $100 billion.
Damodaran’s error, of course, was not recognizing that a great product can expand a market. Uber offered a 10x better offering than taxis: coverage density was higher, which drove down average wait times to under five minutes; geolocation on mobile devices enabled anyone to call a car, nearly from anywhere; payment was done via mobile, meaning customers didn’t need to carry cash; the dual rating system ensured quality; and a digital record of each ride meant that Ubers were safer than cabs.
In New York, the taxi capital of America, ride-hailing apps quickly overtook traditional cabs, revealing that a 10x better product would crowd in more riders.
As Gurley puts it: “The past can be a poor guide for the future if the future offering is materially different than the past.”
Sizing the market for a disruptor based on an incumbent's market is like sizing the car industry off how many horses there were in 1910.Gurley also points to a similar mistake made by McKinsey, forty years ago:
“In 1980, McKinsey & Company was commissioned by AT&T (whose Bell Labs had invented cellular telephony) to forecast cell phone penetration in the U.S. by 2000. The consultant’s prediction, 900,000 subscribers, was less than 1% of the actual figure, 109 million. Based on this legendary mistake, AT&T decided there was not much future to these toys. A decade later, to rejoin the cellular market, AT&T had to acquire McCaw Cellular for $12.6 billion. By 2011, the number of subscribers worldwide had surpassed 5 billion and cellular communication had become an unprecedented technological revolution.”
Great products expand markets. Airbnb popularized home-sharing. Tesla brought electric vehicles mainstream. Red Bull effectively created the energy drink market, which now comprises $53B globally and is expected to grow 7.2% a year through 2027. (For its troubles, Red Bull owns 43% of the market.)
Taylor Swift is a classic story of a great product expanding a market. That same story has played out time and time again in the technology world. From the venture perspective, not investing in a company based on market size (particularly at Seed and Series A) can be a grave mistake.
For the other four business lessons from Taylor, check out Rex’s post.
How to work with designers: Concrete tips on building rapport and mutual respect
As a PM, everything you do involves working and partnering successfully with other people.
Product management is one of the least siloed functions, where we sit at the intersection between Business (executive & other departments), UX (design), and Technology (engineering). This means if we want to be productive and get our job done well, we have to work with people the right way — because there definitely is a wrong way, and that way sours relationships, affects the culture, and can drastically mess with the quality and speed of work.
In a great post by Julie Zhuo (Ex Product Design VP at Meta), she looks at how product managers can build successful working relationships with design partners.
Rifting off, and building on, her advice from 2013, I’ve added a few other tips alongside hers based on my experience.
Talk about your working relationship directly.
Understand each other's thought processes.
Include designers early.
Bring them enough context, and spend time in the problem space.
Discuss timing early.
Trust their expertise.
Create and align on shared goals.
Sweat the details.
👇
1. Talk about your working relationship directly.
The easiest way to figure out how to have the best partnership with your design team? Just go and ask them.
Designers appreciate the opportunity to tell you what works well for them, and what doesn’t. This sets expectations about the roles and responsibilities early on both sides and removes guesswork along the way as you build a relationship.
And a conversation like this is just the starting point. Expect some differing opinions on how you and design should work together — discuss them, compromise where necessary, and ultimately work together to set up a good baseline.
From here, don’t treat it like a one-and-done situation, especially when you’re new to a team and building the relationship up. Schedule time to check in with each other at different intervals to see how things are going, what’s working, what’s not, etc.
2. Understand each other's thought processes.
PMs and designers are both decision-makers, meaning you’ll often find differing opinions on how to approach a problem. But, there are more similarities in your thought processes than there are differences, like focusing on the JTDB and what the customer needs
Usually, these differing points of view often come from something along the lines of:
Designers will tend to think about the experience, the flow, how consistent the components are, how to make it better work in different resolutions, etc, whereas PMs will be thinking about aspects like current business needs, engineering efforts being worked on, product strategy for the next few periods, what’s on the roadmap, resource allocation, etc.
The advice here is to be aware of these motivations designers have, and more importantly, share the context you have with them so they understand your framework for making a decision.
Another useful tactic — when you disagree with a design decision being presented, just be genuinely curious and ask them to walk you through their thought process. Either you learn something new and you change your perspective, or if there really is a flaw, they may uncover it themselves. Being curious always trumps getting annoyed.
3. Include designers early.
PMs benefit from design expertise throughout the entire product development process.
The earlier you bring design partners in, the greater the context they’ll have, and they’ll be able to develop a keener intuition for the problem and what’s best for the customer — helping you both drive higher quality execution faster.
In practice, this usually starts by inviting your designers to the user interviews you’re setting up. This shows that you understand and empathize with their role — which is rooted in understanding the customer — and respect their expertise and opinions to bring them in from the very beginning.
The earlier you bring designers in, the greater the context they’ll have.
Speaking of.
4. Bring them enough context, and spend time in the problem space.
As you set up the foundation for collaboration with designers, it’s super important to make sure there’s alignment on the context problem, vision, and big picture/strategy.
AKA — you want to spend as much time in the discovery and analysis phase with the designers and dedicate time to working within the problem space.
You want to avoid jumping too quickly into trying to find a solution. This can happen either when (1) a PM tries to start brainstorming with designers before they understand the why — which leads to confusion, or worse, (2) when the PM brings them a fleshed-out solution — leading to frustration.
When you’ve set the context up well and designers feel confident moving forward, you’ll feel the natural shift towards designing and discussing a solution.
5. Discuss timing early.
Designers need to know when you’re expecting to present a mockup. This is project management 101, and hashing out timelines and due dates with them is key to a good relationship.
They’re managing their own backlog of design tasks, and when you need designs — whether that’s UX flow charts, wireframes, or a high-fidelity prototype — make sure priorities are communicated and you agree on expected delivery times.
Here’s a simple and reliable template I’ve found handy when it comes to discussing timing. Ask, in this order:
1. “This is what we need overall from a design perspective”
2. “This is the deliverable we need next”
3. “This is, ideally, when we need it by, and why”
4. “What does your workload look like right now? Is this timing doable?”
This approach does a few things:
It shows you’re structured in your approach to managing the project
It isn’t open-ended, as it starts the discussion off by you proposing initial timing
It doesn’t dictate a time, leaving the designers to agree on what’s doable for them
6. Trust their expertise.
Designers are the UI and UX experts on the team. Even if you‘re a Canva wiz, or have a design background, as the PM that overarching responsibility of design isn’t yours, and you should defer to them on decisions around the user experience.
The designers know what it takes to do their work well — and it’s really important to show trust in their process, their creativity, and the time frames they give you.
An important point around trusting their expertise; don’t treat designers like pixel monkeys. This never goes well. Don’t just bring them a wireframe you whipped up and ask them to make it look pretty. Designers want to understand problems and help create the solution — not do the admin of making a high-fidelity version of a spec.
Also, the more senior the designer, the more abstract the problem they should be solving.
7. Create shared goals.
Having common ground makes moving quickly in the right direction much easier — and common ground is best defined within a team as having shared goals.
As the PM, you’re the facilitator of collaboration. So it’s on your shoulders to be clarifying the context by describing problems and goals, and tying it all back to the product strategy.
Fortunately, designers and PMs share many of the same goals. They are, however, often just described differently. For example, while a PM may think in terms of “business goals”, like driving conversions — and a designer in terms of “design goals”, like consistency, accessibility, and usability — these are actually shared goals.
Those examples of design goals are clearly supporting goals towards that overarching business goal of better conversions — if you have a more accessible and user-friendly onboarding flow, conversions go up.
What you want to be doing is establishing and communicating what those shared goals are, and always getting centered around the most important shared goal of them all — focusing on user benefit. You’re both in extremely customer-centric roles, so align on that.
As Julie says, “to speak the language of designers, stop talking about metrics and start talking about users.” This takes us to our last point.
7. Sweat the details.
The most direct path to a designer’s heart is to care about the details.
Seriously, want to send hearts fluttering ablaze with joy and delight? Implement a mock with every pixel in place. Set a high quality bar that doesn’t permit jankiness. Go the extra mile to get a small design detail right. Or spend an extra night building something for the express purpose of delighting a user. Every single designer I know loves to work with engineers and PMs that value design—would gladly give up nights and weekends just to sit together and make stuff happen because everybody believes in it, and everybody on the team just wants to build some really useful, really stellar, really next-level shit.
— Julie Zhuo
To recap:
Talk about your working relationship directly.
Understand each other's thought processes.
Include designers early.
Bring them enough context, and spend time in the problem space.
Discuss timing early.
Trust their expertise.
Create and align on shared goals.
Sweat the details.
If you see a problem, solve it: Adding value in your early career.
Through Substacks epic recommendation engine, I very recently stumbled upon a publication by Ami Vora,
. Ami is the VP of Product @ WhatsApp, and her tech career spans 20 years across companies like Facebook, Instagram, and WhatsApp.I read a few of her posts, and one I liked, in particular, was where she sheds some insight on how people earlier on in their careers can get hands-on and add value.
Simply, if you see an unsolved problem, jump on it. As she explains:
In retrospect, jumping straight into solving whatever problems I saw was a great way to keep adding value. But at the time, I constantly had to remind myself to just take on those problems instead of waiting for someone else to give me permission.
What held me back? I used to think there was an omniscient group of “senior leaders” who would know the most important problems and hand them out to the people who could best solve them. I longed to be the kind of person who got called to solve the hardest company problems, and I waited to get tapped on the shoulder by someone who knew better.
But over time, I realized that if I'm close enough to see a problem, I'm probably best positioned to solve it. In fact, what if no one else can even see it? That realization gave me permission to step into problems, even when I’m not the perfect person to solve them.
One simple way to do this: if I’m new to an area, I ask my manager or an experienced peer a simple question: “What are you doing today that I can take off your plate?” That way they can scan their calendar and pick something tangible and immediate, like running a meeting or tracking down an escalation. They get a little time back, and I can handle something that will quickly expose me to the most important work of the team.
Of course, like with any sort of growth, this route comes with a lot of normal frustrations. When I run into those, it’s helped to:
Remember that stepping up is supposed to feel terrifying. I’m tackling something hard that I’ve never done before – no wonder it’s uncomfortable!
Take perfection off the table. I know I’m not the exact right person to solve most problems. But once I remember there’s no perfect solution to a problem, then I can propose my imperfect solution – and almost always, making progress today is better than waiting for something perfect.
Think of learning as the reward. Most of the time, stepping up doesn’t mean an immediate reward. But every time, I've learned something big – about a new product area, or how to work with different people, or cutting-edge industry info.
Sometimes I’m worried about stepping on people’s toes, but every leader I’ve worked with has been thankful to have extra attention. Every time I’ve said, “I notice you spend a lot of time doing X, so I took a stab at a draft for you. Does this help? Feel free to use it,” a leader has been happy to consider whatever I’ve put together — especially when I'm clear that I'm not looking for credit but just to get to a better solution.
Thinking about new areas like this isn’t easy. How can I justify spending time on added problems when I’m already stretched at my core job? But taking on these sorts of problems is how I gathered skills for future jobs. And making time for these new problems meant my job always stayed fresh and interesting, with new domains and new people I’d never run across otherwise.
For more of Ami’s insights into product, management, leadership, and all the hard things no one tells you about when you’re lucky enough to hit hypergrowth — follow her work at The Hard Parts of Growth
And to wrap up for today, let’s circle back to AI briefly with some hot news that dropped this week.
Tech leaders urge a pause in the 'out-of-control' AI race
A petition has been signed by Elon Musk (who co-founded OpenAI), Steve Wozniak, Andrew Yang, and over 1K other prominent folks in tech, urging for a halt in the training of high-capacity AI models more powerful than GPT-4. They’re calling the pause necessary as we develop safety and ethics protocols for the technology, ensuring we all get to enjoy AI and keep our jobs.
Here’s an excerpt from the open letter.
AI systems with human-competitive intelligence can pose profound risks to society and humanity, as shown by extensive research. and acknowledged by top AI labs. As stated in the widely-endorsed Asilomar AI Principles, Advanced AI could represent a profound change in the history of life on Earth, and should be planned for and managed with commensurate care and resources. Unfortunately, this level of planning and management is not happening, even though recent months have seen AI labs locked in an out-of-control race to develop and deploy ever more powerful digital minds that no one – not even their creators – can understand, predict, or reliably control.
Contemporary AI systems are now becoming human-competitive at general tasks, and we must ask ourselves: Should we let machines flood our information channels with propaganda and untruth? Should we automate away all the jobs, including the fulfilling ones? Should we develop nonhuman minds that might eventually outnumber, outsmart, obsolete and replace us? Should we risk loss of control of our civilization? Such decisions must not be delegated to unelected tech leaders. Powerful AI systems should be developed only once we are confident that their effects will be positive and their risks will be manageable. This confidence must be well justified and increase with the magnitude of a system's potential effects. OpenAI's recent statement regarding artificial general intelligence, states that "At some point, it may be important to get independent review before starting to train future systems, and for the most advanced efforts to agree to limit the rate of growth of compute used for creating new models." We agree. That point is now.
Therefore, we call on all AI labs to immediately pause for at least 6 months the training of AI systems more powerful than GPT-4. This pause should be public and verifiable, and include all key actors. If such a pause cannot be enacted quickly, governments should step in and institute a moratorium.
To be clear, they’re not calling for a total stop in AI development. Rather, a very specific “step back from the dangerous race to ever-larger unpredictable black-box models with emergent capabilities.”
The response?
So far…not a peep from OpenAI or Google.
What do I think?
I tend to agree with Andrew Ng. The goal of this pause (responsible and safe AI) is true and noble. But what they’re suggesting really isn’t feasible. We’ll just see a game theoretic outcome where some companies will continue development, if not simply for fear others might. That’s game theory 101. As Andrew says:
The call for a 6 month moratorium on making AI progress beyond GPT-4 is a terrible idea.
I'm seeing many new applications in education, healthcare, food, ... that'll help many people. Improving GPT-4 will help. Lets balance the huge value AI is creating vs. realistic risks.
There is no realistic way to implement a moratorium and stop all teams from scaling up LLMs, unless governments step in. Having governments pause emerging technologies they don’t understand is anti-competitive, sets a terrible precedent, and is awful innovation policy.
Responsible AI is important, and AI has risks. The popular press narrative that AI companies are running amok shipping unsafe code is not true. The vast majority (sadly, not all) of AI teams take responsible AI and safety seriously. Let's invest more in safety while we advance the technology, rather than stifle progress.
A 6 month moratorium is not a practical proposal. To advance AI safety, regulations around transparency and auditing would be more practical and make a bigger difference.
I’m curious to know what you think about a pause like this. If you have a take, drop it in the comments below.
Related reading: Why Americans fear the AI future, and The AI industry really should slow down a little
And that’s it for today! I’m truly very grateful you give me your time every week. There’s a lot out there competing for your attention, and the fact you choose to spend some of it here—not to sound cheesy—is an honor. So, thank you. 🙏
I hope you have a wonderful weekend, and I’ll see you on Wednesday for our next deep dive. Hint, catching up on Why Vine Died will give you some nice context. ;)
Until then.
— Jaryd ✌️