There are some weeks where I don't read or come across anything too interesting. And there are weeks like this one (week of 2025-05-26) where when it rains, it pours. So apologies in advance for a long post this time round.
Funnily Serious
It takes real talent to present something serious in a funny way. I am not talking about satire here, but presenting actual, useful information in a funny way. And I don't come across such articles that often. But we are talking about this week, and so I have two on hand :-)
Exhibit 1: This article about Pinduoduo and their recent earnings release. Here are some select snippets from that article if you are wondering if it would be worth your time.
Pinduoduo’s management communicates as if they were electing the next Pope.
Think of investing like playing poker: you know the rules, the probabilities, and the deck. The rest is a combination of skill and luck. Pinduoduo, on the other hand, feels like a game where you don’t even know what’s in the deck.
Apparently, these results shell shocked some people. And yes, if you care about these things, the results wildly missed analyst estimates—though I imagine most models were built on the same data foundation as horoscopes.
Link to the full article: Pinduoduo - Game over?
Exhibit 2: This article about making some changes to your Google Docs settings for a more pleasing reading experience. Are you serious? A doc about making updates to Google Doc settings is funny, you ask? Ok, here you go:
Page breaks are for those who either need to use a fax machine or if you’re trying to make an office memo look a little longer than it is.
I will stop here. You can read the rest for yourself: How to take your Google Docs from, "OMG, I'm going blind!" to a Zen like experience.
Bonus: This one is not funny, but personally it is funny and ironic that I stumble on this article on the very day I hit the Publish button after 7 long years! Not sure if this is a good sign!
Product Talk
AR/VR
Here's a fantastic article that talks about the current AR/VR landscape in China. My key takeaways:
China has a vibrant AR ecosystem led by startups and not tech giants (unlike the US where Meta and now Google are taking the lead)
This video!
I learned something about microdisplays and waveguide.
Because of dependence on external computing power, companies are looking to partner with phone companies that control the smartphone ecosystem.
XREAL seems to have an AR glass that looks really impressive, particularly the cost at $600!
RayNeo, owned by TCL, chooses to not use the most advanced tech for the sake of it, and instead focuses on keeping the costs down while providing the AR benefits.
Rokid, INMO and Meizu seem to have integrated AI features into their offerings, while keeping the cost low.
Full Article: China’s Hundred Lens War
Amazon Key
ByteByteGo did a deep dive into how the Amazon Key product came to be, and it is a masterclass in creativity, innovation, and perseverance. Some key highlights:
The original idea was proposed for optimizing usage of underutilized parking lots.
It took 6 months of relentless pitching this idea to get funded by an org within Amazon - only to see no customers sign up for this product after building the product for a year.
At this point, most would have given up, but instead, they pivoted to providing building access for last-mile delivery ops.
In a way I still don't understand, they made this work for 80% of doors and garages.
The initial version required internet access via an ethernet connection. Later, they shifted to cellular connectivity - not one, but two cellular providers, because redundancy.
Use tools where they fit. Even though they migrated away from Lambda, they didn't abandon it entirely because it fit certain tasks perfectly.
Somehow, reading through this article felt like the engineering equivalent of a HBR case study (in a positive way).
Here's the article: How Amazon Key Opens 100 Million Doors Every Year
Brick
A physical brick that can lets you "lock" and "unlock" distracting apps on your phone. To unlock those apps, you need to physically tap the brick on your phone. Neat idea?
This Brick Gave Me My Life Back (From My Phone)
AI Stuff
Everything Wrong with MCP: I have seen quite a few articles on MCP (the USB standard equivalent for AI Agents), but none that talked about any potential problems or challenges with MCP. Well, finally, here's an article that points out the fallacies with MCP, primarily around security and UX.
Microsoft is dogfooding AI dev tools’ future: Articles from Pragmatic Engineer are generally great, and worth your time, and this one is no different. This line around "too much optimism around AI" stuck with me:
if there’s too much pessimism and it leads to inaction, then startups coming from nowhere can become billion-dollar businesses and rivals. This is exactly what happened with Google and Transformers: the search giant invented Transformers in 2017, but didn’t really look into how to turn it into products. So, a small startup called OpenAI created ChatGPT, and now has more than 500 million weekly active users, and a business worth $300B. Google is playing catchup. It was a costly mistake to not be optimistic enough about the impact of Transformers and AI!
It's interesting to see how Microsoft is approaching AI (vs) the other Big Tech companies. Microsoft seems to be going after developer tools big time, and emphasizing that AI is working for devs, and not replacing them. And they are taking the pain to emphasize this in the UX.
AI Horseless Carriages - Another fantastic essay on how today's AI apps are badly designed. And his illustration of the problem with Gemini on Gmail, and his solution using System Prompts vs User Prompts is worth reading. Here's my highlight excerpt from that essay:
The "old world thinking" that gave us the original horseless carriage was swapping a horse out for an engine without redesigning the vehicle to handle higher speeds. What is the old world thinking constraining these AI apps?
The Great Unbundling of Work - A very different take on "AI is going to take over our jobs" narrative. The author argues that chunks of our expertise become commoditized going forward, and we need to figure out what we are going to fill up those chunks with.
AI has cracked the outer shell of most knowledge work. What remains is the messy, judgment-heavy, human-centre and is harder. It's also more valuable - and even more so as we step forward in this age. But the crisis that no one discusses is that your identity was wrapped up in those tasks. The ones now being automated. The ones that made you feel like an expert. The jobs aren't disappearing. They're being hollowed out. And the question becomes: what are you filling that space with?
Autonomous Driving
If you are ever curious about autonomous driving and robotaxis, and how that might play out in reality, I recommend reading Phil Koopman's article below.
All Robotaxis Have Remote Drivers
On China
Of late, I am reading two different perspectives. One is that China is winning. The other is that China is facing a crisis. Both can't be true though, right? Well, two articles outline how that is a possibility.
Article 1: The "China as superpower" vs "China in crisis" narrative
Article 2: Lets talk about bipolar disorder...in China watching
Phew... we made it to the end. Happy reading (if you choose to). Until next time :-)