Today’s Highlights

• AI: AgiBot’s Game-Changing Robotics Dataset: Smarter, More Capable Bots in Homes and Workplaces

• Design: The Year in Magazine Covers: Who says print is dead?

• Science & Tech: Life Under the Sea: Deep is Building Modular Underwater Homes for Scientists

• Founding: 3 Lessons from 3 Struggling Startups 

• Product: What Turns Your Newbie Users Into Loyal Devotees? : Finding the “Aha” Activation Metric

• Today’s AI image: AgiBot’s Feast For AI

• Quote for the day: Who’s Truly Rich?

AI

AgiBot’s Game-Changing Robotics Dataset: Smarter, More Capable Bots in Homes and Workplaces

Chinese robotics powerhouse AgiBot has launched AgiBot World Alpha, a groundbreaking open-source dataset designed to elevate the training of general-purpose humanoid robots.

All You Need To Know

The dataset includes over 1 million robot trajectories collected from 100 robots tackling a variety of tasks in industrial, domestic, and commercial environments.

Tasks range from simple object handling to intricate multi-robot coordination, with a significant 40% dedicated to household activities.

A Dataset Like No Other

AgiBot World Alpha is a behemoth compared to existing datasets, boasting 10x more navigational data than Google’s Open X-Embodiment and spanning 100 times as many scenarios.

Researchers and developers can grab the entire dataset for free on platforms like HuggingFace and GitHub, levelling the robotics playing field.

Why is this a big deal?

Robotics innovation has long been hampered by the lack of top-notch training data.

This release not only fuels the development of smarter household robots but also gives smaller developers a fair shot at building cutting-edge solutions.

In short, AgiBot’s massive contribution could be the jolt the robotics world has been waiting for.

Read More

Also in AI

• Hugging Face’s Featherweight AI Framework: A minimalist Python library to create AI agents without the coding headache

• AI leadership roles hit record highs in 2024—organisations are officially all-in on AI

• OpenAI’s new safety protocols are teaching AI to "think twice" before responding

• Leeds researchers use AI to spot undiagnosed heart issues early

• AI hallucinations are helping scientists make groundbreaking discoveries—who knew daydreaming bots could be so productive?

• Alibaba slashes cloud AI costs by up to 85%

• A voice-based AI app from the University of Toronto could replace your blood pressure cuff

Design

The Year in Magazine Covers: Who says print is dead?

Print magazines, though grappling with declining sales, down 2.4% annually since 2019 (IBIS World, remain culturally significant.

Despite the digital era, their covers flood online spaces, acting as cultural snapshots and visual timelines of major events.

Cultural and Political Turmoil in Design

Designers leaned into protest art, with Trump-centric covers dominating yet evolving in their approach.

Jaap Biemans, art director of Volkskrant Magazine, suggests future designs may shift focus from Trump’s image to his policies for fresh takes.

The fall of Joe Biden, Kamala Harris’s fleeting rise, and global crises like Gaza, Ukraine, and the Olympics also shaped cover narratives.

Taylor Swift’s Eras Tour and its cultural impact graced the pages as well.

Outstanding Covers of 2024

• Bloomberg Businessweek: A January maze cover brilliantly foreshadowed a chaotic year.

• Politico: Delivered a fresh spin on Trump with a concept-driven design.

• The New York Times Magazine: Blurred lines between past and present with haunting historical visuals.

• The Economist: Used black space masterfully in a sombre portrayal of Ukraine’s President Zelenskyy.

• TIME: Ingeniously depicted Biden’s exit and Harris’s rise with two consecutive, thematically linked covers.

Why is this a big deal?

Magazine covers continue to serve as powerful cultural artefacts—merging art, commentary, and storytelling.

They reflect the zeitgeist, offering both critique and solace in a world that often feels overwhelming.

So, while print may be shrinking, its impact remains larger than life.

Read More

Also in Design

• YouTube's 'Play Something' Button: YouTube is testing a floating button that randomly selects videos via the Shorts player

• Samsung One UI 7’s Task Manager Revamp: Offers an intuitive Task Manager for better multitasking and app control

• The Psychology of Fonts: Serif equals tradition, sans-serif screams modernity, and script whispers creativity. Choose wisely.

• Diary Studies for Better UX: A bank’s redesign of an ancient loan tool used diary studies to unearth real-world pain points

• Ansible on IBM Z: Spotlight on Stewart Francis

• 25 Predictions for AI and Creativity in 2025: Orchestration and judgment are the new must-have skills

• Hypothesis-Driven Design: Tackle complex problems by framing user-centric hypotheses

Science & Tech

Deep is Building Modular Underwater Homes for Scientists

A British startup, Deep, is crafting a bold vision of life beneath the waves, starting with trials in an old quarry straddling Wales and England.

By 2030, they aim to establish a permanent human presence underwater, reaching depths of up to 200 metres.

From Vanguard to Sentinel: The Habitat Evolution

• Vanguard: A compact, transportable underwater home for three divers, capable of week-long missions. Testing kicks off in early 2025.

• Sentinel: The ambitious follow-up, scheduled for 2027, will feature modular pods housing up to six people, complete with wet labs and private quarters. It’s designed for extended stays of up to a month, with resupplies every 28 days.

Why Bother Living Underwater?

Diving at 200 metres currently offers just 10 minutes of work time, followed by hours of decompression.

With habitats like Sentinel, researchers can achieve seven years’ worth of work in just 30 days.

Over 90% of marine biodiversity resides within 200 metres of depth, yet we’ve only explored about 20% of it.

Understanding these ecosystems is vital for tackling climate change, as oceans absorb a quarter of human-caused CO₂ and most of our excess heat.

A Nod from the Experts

Skeptics aside, the project has earned praise from industry heavyweights like John Clarke, former lead scientist of the U.S. Navy Experimental Diving Unit.

He lauds Deep’s innovative approach and their strides in oceanic exploration.

The ocean’s depths are calling, and with Deep’s efforts, it seems we might just answer—not with snorkels, but with homes.

Read More

Also in Science & Tech

• Inside Apple’s Secret iPhone Camera Labs: Better slow-mo, customisable audio sliders, and a billion pixels per second

• o3, OpenAI’s Latest Brainchild: Smashing benchmarks and giving competitive programmers a run for their code

• World’s Oldest Cheese Unearthed: Turns out, ancient humans had a taste for dairy—though we might skip this slice

• Do Organs Hold Memories? A Mind-Bending Read

Founding

3 Lessons from 3 Struggling Startups 

In this blog post, an angel investor talks about how three of his startup investments turned out to be challenging and why each of them failed. Let’s have a look.

Startup X: The Fast-Growth Pitfall

As the writer says, this company was growing rapidly, almost 20% month over month. However, growth stalled, and revenue began to decline.

Why? 

It was simple. The company managed well to allure new customers with a great product in the beginning, resulting in swift graphs. 

But overtime, it couldn’t satisfy customers’ needs sufficiently that raised the churn rate as well as doubt within potential buyers. 

Startup Y: Breathless Cash Reserves

A tough market and limited cash reserves proved too much for the founder to overcome. Despite strong growth, the company couldn’t achieve the necessary traction and ultimately ran out of cash.

Like Startup X, he invested between rounds, and as the company burned through cash, the lack of a financial buffer became critical.

Startup Z: Icy Co-founder Relationship

Initially, this company gave off surefire success vibes with an amazing product and rapid growth. 

However, a co-founder dispute led to one leaving, and the company’s fortunes quickly declined.

This was the hardest loss to predict for the investor. A strong co-founder relationship is crucial. Cofounders are pillars dependent on each other to hold the organisation. One breaks and the organisation collapses.

Read more

Product

What Turns Your Newbie Users Into Loyal Devotees? : Finding the “Aha” Activation Metric 

To work out your product’s activation metric, answer this one question: What actions of your user tells you that they use your product regularly and religiously?

PostHog, an open-source product OS, figured that the users who watched 5 or more session replays and set a replay filter turned out to be their most loyal ones. Here’s

How YOU Can Define Your Activation Metric

#1 Identify where your user starts engaging: Select events that represent the start of user engagement.

#2 Then identify events when the user goes a step further: Choose events that might signal increased engagement (e.g., key actions within the product).

#3 How many times the event must be performed to tag the user a “hit”: Define what constitutes a successful user by specifying the number of times each event must be performed.

#4 Analyse extended engagement of the “hit” users. Like months later. Track whether these successful users continue to use the product months later.

#5 Repeat the process: Test different event groups to find the most reliable activation metric.

Read more

Today’s AI Image

AgiBot’s Feast For AI

Quote of the Day

Who’s Truly Rich?

"We are not rich by what we possess but by what we can do without."

Immanuel Kant

What we’re working on

🧞Your wish is my command