Martin Schmalz20 February 2020

Artificial Intelligence, Big Data & China - Martin Schmalz - #144

0Frameworks

9Insights

Insights & moments

The myth-busts, hot takes, explainers, and tools worth keeping.

Myth Buster· 1

Myth Buster07:30

AI Doesn't Think — It Just Predicts

Contrary to popular belief, most AI today isn't about creating thinking machines. It's about using large datasets to make faster, cheaper, and more accurate predictions. AI excels at generic prediction tasks — like ad targeting or loan default risk — but it cannot think, reason, or innovate like humans. The idea that AI will replace all human jobs is largely exaggerated.

AI is not about artificial intelligence in the human sense.
It's primarily used for statistical prediction from large datasets.
Computers don't think — they compute based on patterns.
Human judgment is still essential for non-generic predictions.

“Computers don't think. Like artificial intelligence has very little to do with intelligence in a broader sense.”
Martin Schmalz · 09:30

#ai myth#machine learning#prediction#human vs machine

Hot Take· 1

Hot Take12:00

Top Executives Are Irreplaceable Decision Engines

The most valuable executives aren't just leaders — they're complex decision-making systems that synthesize vast, disparate information to make strategic calls. Unlike AI, they can predict outcomes in novel situations — like launching a phone without a keypad — because they combine intuition, creativity, and experience in ways machines can't replicate.

Top executives make decisions AI cannot predict.
They synthesize complex, non-quantifiable variables.
Creativity and intuition are uniquely human advantages.
AI can't predict first-time market disruptions.

“The real value that is added by Tim Cook at Apple or by Elon Musk... is their ability to compile a lot of disparate complex…”
Host · 12:30

#executive leadership#decision making#ai limitations#creativity

Explainer· 4

Explainer00:00

Why Sleeping in Two Places Predicts Loan Default

Location data can reveal surprising insights about financial risk. If someone sleeps in two different locations interchangeably, it correlates strongly with higher credit risk. This pattern may indicate personal instability, such as having a secret relationship that could lead to costly divorce. While the exact reason isn't certain, the data shows a clear predictive link to future loan default.

People who sleep in two locations are more likely to default on loans.
This pattern may signal personal instability or hidden relationships.
Location data is highly effective for predicting financial behavior.
The exact cause isn't proven, but the correlation is strong.

“If people start sleeping in two different locations interchangeably at night, that tends to be a really bad credit risk.”
Martin Schmalz · 00:00

#big data#credit risk#location tracking#predictive analytics

Explainer15:30

How Fast You Type Predicts Financial Risk

The speed at which someone fills out an online form can be used to assess their creditworthiness or fraud risk. Slow typing or frequent typos may indicate low intelligence, carelessness, or fraudulent intent. Chinese tech firms already use this data point, showing how behavioral micro-indicators are being leveraged to predict real-world outcomes like loan repayment.

Slow typing on forms may signal fraud or low intelligence.
Typos can indicate carelessness or higher insurance risk.
Behavioral data like typing speed is now used in risk modeling.
China is ahead in using such granular data for financial predictions.

“If you can't fill in your social security number real fast, then you're probably a fraud or I don't know, not particularly intelligent or something.”
Martin Schmalz · 16:30

#behavioral data#fraud detection#typing speed#ai prediction

Explainer26:30

Why China Is Ahead in AI and Big Data

China leads in AI adoption due to a combination of massive population data, fewer privacy regulations, integrated 'super apps' like WeChat, and strong government and corporate investment. With fewer legal barriers like GDPR, Chinese companies can combine data across services — from messaging to payments — enabling richer behavioral predictions and faster innovation.

China's large, homogeneous population provides vast training data.
Fewer privacy laws allow broader data collection and use.
Super apps like WeChat integrate multiple services, enriching data.
Companies like Ping An employ thousands of AI engineers.

“China has a large population in a reasonably homogeneous economic system... you have a lot of data.”
Martin Schmalz · 26:30

#china#ai leadership#big data#super apps

Explainer36:30

How Privacy Became an Antitrust Issue

Regulators now treat privacy as a form of competition. In Europe, Facebook was blocked from merging WhatsApp and Facebook data because offering better privacy could be a competitive advantage. By not allowing users a privacy-respecting alternative, dominant firms abuse their market position — a new frontier in antitrust law.

Privacy is now a competition issue, not just ethics.
Germany's Bundeskartellamt ruled Facebook abuses its dominance.
Lack of privacy-respecting alternatives harms consumers.
This sets a precedent for regulating data monopolies.

“People care about privacy. If there was a social media network site similar to Facebook that offers a similar benefit, but that actually cares about…”
Martin Schmalz · 37:30

#antitrust#privacy#facebook#regulation

Story· 1

Story46:00

How People Game Fitness Trackers and Why It Matters

When insurance companies offer discounts for hitting step goals, people find creative ways to cheat — like strapping Fitbits to dogs or buying motorized devices to fake activity. This shows that data collection systems are part of a dynamic game between users and companies. Detecting such behavior requires human insight, not just algorithms.

People cheat fitness trackers to get insurance discounts.
Devices exist to automatically shake Fitbits all day.
This creates a cat-and-mouse game between users and companies.
Humans, not AI, detect and respond to such behavior.

“In China you can actually buy a small electrical device that does nothing else than shake around your Fitbit during the day on your desk.”
Martin Schmalz · 47:00

#fitness trackers#data gaming#insurance#human insight

Tool· 1

Tool22:00

Using Email Domains to Infer Customer Age

Insurance and financial companies use email domains like AOL or Yahoo as a proxy for customer age. Since older users are more likely to have these legacy addresses, companies can infer demographic trends and tailor products accordingly. This is an example of how seemingly minor data points can inform business strategy.

AOL and Yahoo email addresses correlate with older users.
Companies use this to target or exclude certain demographics.
It's not about the email provider, but what it signals.
Such proxies help build risk profiles from limited data.

“Being an AOL user correlates with characteristics or features of people that might very much matter for the riskiness of your insurance portfolio.”
Martin Schmalz · 22:30

#email data#demographics#insurance#data proxies

Takeaway· 1

Takeaway31:00

Privacy vs. Convenience: The Real Tradeoff

Most people accept data collection because the convenience outweighs privacy concerns. From facial recognition at airports to cashless payments in China, users trade personal data for speed and ease. The real issue isn't whether data is collected — it's whether users are aware of it and feel they have a choice.

People accept data tracking when it makes life easier.
China's convenience-driven model shows this tradeoff clearly.
Privacy concerns are real, but often overridden by utility.
Transparency and user awareness are key to trust.

“It's just so much more convenient. And this is basically where things come down on.”
Martin Schmalz · 32:00

#privacy#convenience#data ethics#user awareness