Session 02


Enjoy Boston! 

1.

 
Read and liked: This passage from the link you sent. A few thoughts: 
  • This is what I will focus on for the designs over the next few days (trying to figure out the unique interaction paradigm between people : product reviews  // people : businesses (longer term) 
  • Should Jordan be on our investor list? 

Anyway, these are just two examples that I think highlight a battleground where many more AI applications will distinguish themselves from the field. Each use case that contemplates autonomy calls for a different interaction paradigm between a user and automation, and those that elegantly enable the baton of control to be passed back and forth seamlessly within a given use case will thrive. If you are a product designer and this line of thinking resonates, I’d love to try your product, join your testflight, maybe invest… I’m jordan@pacecapital.com

*Disclosure: I’m not an investor in Granola (thought I wish I was), I am an investor in Tesla
Also like this “Great link to browsingmode, it made me think “we should really just start with a re-skin of chatGPT to make it editorial layout” and have that be a container for testing our ideas“
From a design POV its a good place to start, although I think we need more structure than just chatGPT -- we need some new form factor which may be the “product review page.” 


2. 




  • This was very interesting. I keep coming back to observable behavior, which for me is the “100 open tabs” and messy, fragmented “proof of work” that comes from researching and understanding options, specs, info, and reviews on what to buy, where to go, essentially the “spray” of internet surfing. 

  • Big picture, I think we’re building an AI tool for internet curators / tastemakers / brands. We leverage the coldstart problem by using AI chat and scraping known information about product reviews but then later evolve into first-party authored content that makes it much easier to expand and update the review knowledge graph. 
    •    - reviews, guides, and curations.


3. 


On you quoting Ilya “You don’t have a strong reason to be a founder,” -- I feel like this is where I would disagree with him. I get the point he’s trying to make. But Mark Z didn’t have a need to be a founder anyway. Many people just follow their intution and interest. 

To me, the idea does matter -- in the sense that I know it will adapt and change, within a gravitational bounds set by the inertia of the character of the original idea. I also beleive strongly in founder market fit. 

On starting with product reviews and ultimately lattering up to creating the interaction layer between brands and consumers through AI: I think the anchor point of this idea is right for us and that is why it matters. I’m not going to be motivated to work on some HR tool for 5 years, but creating exceptional and useful AI powered interfaces -- hell yeah. Back to this framing again

“Each use case that contemplates autonomy calls for a different interaction paradigm between a user and automation, and those that elegantly enable the baton of control to be passed back and forth seamlessly within a given use case will thrive.“

This feels like one of the most interesting design problems in the world to solve.


4.

Agree 100% with what you said: “What matters is putting things into contact with people and reality and testing a bunch of things to figure out what sticks.“

 -- and --
“We aren’t going to get confidence from the idea, we’ll get confidence from each other, from converting ideas into reality, from testing out the product and use cases and users (not from a metrics POV but qualitatively).”
On the latter -- my rephrasing of this is “we aren’t going to get confidence to KEEP GOING from the idea alone, but rather our ability to continue converting ideas to reality against that vision.” Ideas give you the confidence to start in earnest. The idea is a servant to the mission, and the mission here is what is deeply compelling to me. 


5.


Both portentially a real tehcnical insight (not sure yet), but also a “game” playing post from Bret Taylor again -- bending the public opinion towards Sierra. 
https://www.linkedin.com/posts/brettaylor_%3F-bench-benchmarking-ai-agents-for-the-activity-7209570429093515264-ivtq?utm_source=share&utm_medium=member_ios


6.


I sat in on Woven’s pitch deck review session today. Have a bunch of good takeaways for us that I will work into our deck -- won’t post photos of the deck here due to courtesy of confidentiality, but I can speak to the high level points I think are interesting for us to noodle on. I think there is a lot of parallel between us and them (or travel apps) in the sense that we’re a similar machine just operating in a different vertical. Both serving up for you content, scraping to create “canoncial pages”: theirs is a canonical Point of Interest Card which includes personalized details on it that matter to YOU as opposed to a generic POI card from Google. For example it will say “we think you’ll like x because of y.” 

Specifically what I thought was interesting was how they framed their opportunity slide 
“Organically acquire loyal customers upstream of booking with a 10x better product consumers love” 

and their biz model slide 
“This new model optimizes for consumers bringing them more deals, aggregates demand, and lowers our cost structure.” 

They make a point to talk about both the business problem and the customer problem (so they anchor the deck with a b2b problem, but then transition the rest of the deck to focus on how they are going to solve the consumer issue and win there. In doing so they demonstrate understanding of the landscape and that theres a real business issue in travel where companies (hotels, booking.com, etc.) need to re-acquire customers again and again, leading to very high CAC. They can change this dramatically by building a sticky, high affinity consumer experience (this also raises LTV). 



7.


When doing a querey for “best travel baby stroller” google shows this -- representing the need to go and open 9 tabs and read all of them. Pain in the ass! Canopy should have read all of these, and more, provide a “synthesis” of top choices by different rating scales / heuristics found in these writeups, and create canonical product details pages for each product reviewed. 



Also check this out -- feels like we could be 10x better than all of these types of sites. Low hanging fruit. Canopy award winners? lol. 


8. 


Starting to sketch. A few very loose ideas: 
  • If we are starting very simple, we can scrape the web as per no.7 above and create our own “guides.” These are going to be 10x better to look at, save you a ton of time, and be updatable. These are “molecules.” We can also start to surface products you might like that are well-reviewed. These are “atoms.” 
  • Tabs might be “for you” you home page, “categories,” “guides,” and “collections” which are things you can save. 
  • Was inspired by the Browsing Mode Aesthetic and so sketched in minimalist light mode. 
  • This is for later -- but the way we show things in this app should be VERY visual -- pinterest x tiktok. Rather than a blog vibe, guides should feel more like a visual reel that cycle through the images from the guide. Then there’s probably cool ways to stop the video to see more or click in a level deeper. Then we could explore gen AI created guide images .. like for the stroller one, it could literally be an AI generated animation or video with the stroller images or even an illustration or something.  
  • Random thought: I LOVE the name Canopy. But out of a whim, what if the app name was “Rabbit” -- inspired by saving you time and visual rabbit holing? It’s very memorable and has a potentially iconic visual brand representation. There could be fun scrolling and loading interactions too :) We could be “Canopy Labs, Inc.” and our product could be something diff. Just a totally random thought.


(update a few hours later ...) Just realizing that this may be too close to Rabbit R1, but maybe not. \(^__^)/