TWS #003: Drones solving the last-mile, AI proves itself in the workplace, Why Robotaxis will work, Apple's Siri screw up
and much more...
Every week, get the latest curated high signals on news, insights, and ideas around technology, science, and business to help you become a better well-rounded thinker and builder. You’ll also get an essay or a deep dive every other week.
Did you know...
In 1994, Masahiro Hara at Denso Wave, a Japanese auto parts company, invented the QR code to track vehicle components more efficiently than traditional barcodes, which couldn’t hold enough data. The QR code’s innovation lay in its 2D matrix design, capable of storing up to 7,000 characters, including Kanji (Japanese writing), and being scannable in any direction. Denso Wave made the tech freely available, leading to its widespread adoption in business marketing by the 2000s—like scanning QR codes on ads, menus, or packaging to access websites instantly. Today, QR codes are a global standard, with over 82% of consumers aged 18-44 using them for business info, all thanks to a car parts problem!Here’s this week’s scoop:
Drones can finally solve the last-mile problem
AI proves it can cut costs and increase revenue for orgs
How Apple screwed up its Siri makeover (finger-pointing)
Why China is abandoning its data centers
The economics of Robotaxis
Explaining AI Agents
and much more…
Let’s dive in.
🔥 Nuggets for the Road
MCP: The new “USB-C for AI” that’s bringing rivals together [LINK]
“I am rich and have no idea what to do with my life” — ex-Loom founder, acquired. by Atlassian [LINK]
Prep or Perish: Mastering In-the-Moment Communication — a three-part miniseries about keeping your cool when speaking on the spot. [LINK]
Building an AI Agent from scratch — fun little weekend project [AMP]
Michael Seibel (YC Partner Emeritus) on 5 lessons from 20 years in startups
📰 The Signal
Drone startup Zipline has finally entered the delivery space.
Zipline, a drone delivery pioneer, has partnered with Walmart to expand its services to Dallas, Texas. Using their P2 Zips drones, the company delivers packages up to 8 pounds (4kg) within a 10-mile radius in 30 minutes, with precision landing and tech using Nvidia chips for obstacle avoidance. These drones operate in winds up to 45 mph. Walmart, with 4,600 stores within 10 miles of 90% of the U.S. population, has completed over 120,000 drone deliveries since 2021. Zipline, valued at $4.2 billion in 2023, has logged 100 million autonomous flight miles and 1.5 million deliveries since 2016. All of these “training miles” were performed in emergency medical supply deliveries across Africa. Drone delivery is now transitioning from niche to mainstream, with faster, cheaper logistics. This partnership signals the embrace of autonomous tech, with the potential to disrupt last-mile delivery. [LINK]
The 2025 AI Index Report (Stanford University)
The Stanford AI Index Report 2025 tracks AI’s global progress. AI models improved significantly in 2024. AI is now pretty much everywhere, with 223 FDA-approved medical devices in 2023 and autonomous services like Waymo’s 150,000 weekly rides. The U.S. produced 40 notable models, China 15, and Europe 3, though China nears parity in quality. Global optimism for AI grew, with 83% in China and 39% in the U.S. seeing benefits. But here’s the graph that really got my attention: AI is so useful that across all typical business units i.e., marketing, HR, engineering etc. there has been a proven cost decreases (left side) but also strong revenue increase (right side) for particular areas like marketing/sales, engineering, and operations. It’s just a matter of time before AI begins to realize revenue for compliance, HR, and IT. [LINK]
Why China really abandoned its Data Centers
ICYMI, China built over 500 AI data centers from 2023 to 2024, with 150 completed by year-end, but 80% remain unused due to weak demand. Financial struggles have led to surplus GPU sales, with Nvidia H100 rental prices dropping from 180,000 to 75,000 yuan monthly. The reason this is happening is because of a shift to inference, driven by models like DeepSeek’s R1 and OpenAI’s o3, which have made many centers suboptimal, causing energy waste. This overbuild reflects the risks of hype-driven investment. Repurposing these centers or optimizing AI workflows might create opportunities. [LINK]
“It’s paradoxical—China faces the highest acquisition costs for Nvidia chips, yet GPU leasing prices are extraordinarily low,” Li says. There’s an oversupply of computational power, especially in central and west China, but at the same time, there’s a shortage of cutting-edge chips.
How Apple screwed up its Siri’s AI Makeover
A lot of internal finger-pointing at Apple's fake demo at WWDC. Last year at WWDC 2024, Apple showcased a revamped Siri with AI-driven features, such as retrieving flight details from emails. These demos were later exposed as fictitious, revealing significant issues in Apple’s AI strategy. The AI/ML group, led by John Giannandrea and nicknamed “AIMLess,” faced criticism for poor leadership. The Siri team prioritized minor improvements, like removing “hey” from “hey Siri,” over transformative changes. They underestimated competitors like ChatGPT. Technical pivots, such as abandoning dual models (“Mini Mouse” and “Mighty Mouse”) for a single cloud-based model, led to engineer frustration and departures. Don’t think Steve would have let this fly. Here’s the summarized version from the REDDIT post so you can read it here. [LINK, REDDIT, MACRUMORS]
Apple India Produces $22 Billion of iPhones in Shift From China
Apple produced $22 billion worth of iPhones in India in the fiscal year ending March 2025, a 60% increase, accounting for 20% of global output. Exports reached $17.4 billion, produced by Foxconn and Tata Group facilities. This shift from China, due to geopolitical tensions and COVID-19 disruptions, positions India as a key manufacturing hub. Apple’s move clearly shows India’s rising role in global tech manufacturing. It could attract more companies, spurring investment in supply chain tech and logistics. Challenges like infrastructure remain, but this trend signals a reshaping of global trade dynamics. [LINK]
The Economics of Robotaxis (and why it works!)
Waymo and Tesla are leading the robotaxi revolution, with Waymo holding 25% of San Francisco’s market and planning expansion to 10 cities by 2025. Tesla aims for an Austin launch this summer and California by late 2025. Robotaxis cost $1.1/mile vs. $2 for Uber/Lyft and have 90% fewer accidents. The global market could hit $5 trillion, with Tesla’s data and integration advantages poised for dominance in the future. Robotaxis could redefine transportation, reducing costs and accidents while challenging car ownership. The $5 trillion market offers huge amounts of opportunities in autonomous tech and mobility services. [TOMAS PUEYO]
AI Agent Breakthrough (by YC)
A useful video if you want to explore how AI agents work at a basic level. There is a strong focus on Manus AI, which is a general-purpose AI agent platform that has developed a unique and powerful way to create and operate Agents. It dives into how Manus’ architecture sets it apart, what it says about the broader debate around GPT wrappers, and why they might actually be more valuable than they first appear. [YOUTUBE]
A Science-based Guide to Thinking Creatively—With LLMs
This one explores how LLMs enhance creativity. LLMs shorten feedback loops, generate ideas, and articulate taste, acting as research assistants or metaphor generators. They complement, not replace, human creativity. LLMs as creative tools open new possibilities for writers, artists, and researchers. Startups building LLM-powered creative platforms could thrive. But human judgment remains important, ensuring AI augments rather than overshadows creativity. I also created a video on how to write creative prompts. [LINK]
“I sometimes prompt ChatGPT to re-explain the core argument of one of my pieces in the voice of a retired philosophy professor or sci-fi novelist, surfacing analogies or interpretations I wouldn’t have thought of on my own. Another way to practice this kind of thinking is by prompting the model to challenge your assumptions—for example, by asking it to present an opposing perspective, or a critique from a voice unlike your own.”
🧐 Trivia of the Week
Last week’s question: What was the first Graphical User Interface (GUI)?
Answer: Xerox PARC - they did this with the Alto computer in 1973, introducing icons, windows, and a mouse, influencing later GUIs. Apple's Macintosh (1984) and Microsoft's Windows (1985) followed, while IBM still focused on command-line interfaces (LOL!).This week’s trivia:
The answer will be provided in next week’s newsletter.
📝 Essay: Asymmetric Risk
We all know that risk-taking is not everyone’s cup of tea.
In fact, in many cultures and socio-economic environments, the idea of taking any degree of risk is considered taboo and perhaps perceived as a vice.
But here’s the problem.
If no one ever took risks, then we wouldn’t have the many luxuries and experiences we take for granted today. I won’t be exhaustive in providing a list of examples to make my point, but I’m sure you get the idea of what our world could look like if no one took risks.
In general, civilization progresses and advances through a healthy dose of asymmetric risk-taking.
In a nutshell, taking asymmetric risk is about taking/making bets on situations, circumstances, and opportunities with a significant probability of potential upside relative to minimizing the potential downside. Mind you, the reverse can also be true; massive downside versus small upside, but the concept is identical in opposite directions. For most of us, we tend to trend towards the former, so we’ll stick with that.
One perfect example is investing — the most you could lose is the initial outlay on your investment, but the potential gains have unlimited upside.
In principle, the concept of asymmetric risk is straightforward but more challenging to implement in practice.
Here’s how I like to approach it;
Be on the lookout — always be searching for opportunities. They appear everywhere, and it just takes a few moments of your day of focus to figure out if there’s something there worth pursuing. If so, double down; if not, then move on with your day and keep searching.
Be logical and prepared— if you have discovered something worth pursuing, then it’s time to put it on paper and really think about the pros/cons of this opportunity. Every decision has risks involved, so if you’re better able to capture those downsides and upsides, then it puts you in a better position to make a more informed decision with the information you have at that given moment.
Be honest with yourself — be completely honest with yourself. It’s easy for us to get carried away by overestimating the upsides and underestimating the downsides.
Take action — in the end, if what you discover is that the upsides far outweigh the downsides, then perhaps it’s something worth considering. If so, then all you need to do is take action. All of this would have been for nothing unless you took actionable steps. Remember that it’s easy to dream. You need to act to make this a reality.
At the end of the day, risk-taking is part of life. It’s what gets us out of our comfort zone and allows us to grow.
FULL POST BELOW 👇
🙏🏼 Connect with Me
Are you new to the newsletter? Subscribe Here
Learn more about me on my website
Check out my YouTube channel (and subscribe!)
If you’re a founder, apply here (Metagrove Ventures) for startup funding or contact me directly at barry@metagrove.vc
If you think this could be helpful and informative to others, please share it :)
Thanks for reading.
Barry.