View profile

DataScan: Issue #43

Revue
 
Who will win the race to become the world's first AI superpower? Do AI developers need a data marketp
 

DataScan

August 6 · Issue #43 · View online
Curated digest on the world of data.

Who will win the race to become the world’s first AI superpower? Do AI developers need a data marketplace? How did data become a new medium for artists?
💌  Forwarded this email? Join my weekly data digest.

Who will win the race to become the world’s first AI superpower🤖 Reporting for The Verge, James Vincent provides an interesting commentary on the current AI race between China and the US and explains how China has “all the ingredients it needs to move into first” place. 👑  Not only is China “significantly more permissive when it comes to users’ privacy”, it also has “strength in numbers”:
To build great AI, you need data, and nothing produces data quite like humans. This mean’s China’s massive 1.4 billion population (including some 730 million internet users) might be its biggest advantage. 
Or will one of the tech giants beat them to the punch? 🤔
Monica Rogati’s AI Hierarchy of Needs points out that collecting high-quality data is vital to provide the “solid foundation” needed to reach the pinnacle and so succeed in AI and deep learning:
The Data Science Hierarchy of Needs, by Monica Rogati
KEY TAKEAWAY - Rogati highlights the overwhelming noise around the “current AI hype” and how everyone wants to get involved - but if an AI is fed garbage, the output will be the same:
People try to plug in data that’s dirty & full of gaps, that spans years while changing in format and meaning, that’s not understood yet, that’s structured in ways that don’t make sense, and (then) expect those tools to magically handle it
So do AI developers need a marketplace to get better data? 🛍  Philip Keys, from IoT data marketplace provider Intertrust, reiterates that how “without data, AI ends up as an empty vessel, a group of algorithms churning without producing anything of interest”. Keys references the growing trend of “data monetisation” or “data marketplaces” - which offers AI researchers (or “at least those whose organisations have the cash to pay for it”) the opportunity to access previously unavailable anonymised data. 📊
– Keys predicts that a large proportion of data on data marketplaces will be from IoT devices. 💡 However, Carlos Ariza, BI & Analytics Expert at PA Consulting Group, points out that the best next-gen analytics platforms will be “cloud-based, collaborative, and multi-entity” and - crucially - will “aggregate data from inside and outside sources”. This suggests the need for a wide range of data types to be available in these marketplaces. 🚀 
However, as revealed earlier in the week at Def Con, identifying individuals from anonymised browsing data is “trivial” - raising some serious privacy questions. 😱  Under the false pretence of needing to test an AI advertising platform, German researchers were able to obtain anonymised browsing data on three million users from a data broker for free. There was then no restrictions on how the data could be queried, so they could quickly re-identify “anyone who has a public social media presence”:
It requires an astonishingly small amount of browsing information to identify an individual out of an anonymous dataset of 3 million people. Since everyone’s browsing habits are unique, it only takes about 10 website visits to create a “fingerprint” for an individual based on which websites they are visiting and when.
Miscellaneous
💰  Reddit raised $200 million in funding - now valued at $1.8 billion. 
📈  Snapchat struck a new data deal to equip marketers with more 3rd party measurement. 
🔍  HN hiring trends - most popular programming languages of July ‘17.
🎨  How data became a new medium for artists:
Thousand Little Brothers, 2014. Detail from a composite image made of 32,000 images taken from the artist’s daily life. (Courtesy of Hasan Elahi and Open Society Foundations)
Did you enjoy this issue?
If you don't want these updates anymore, please unsubscribe here
If you were forwarded this newsletter and you like it, you can subscribe here
Powered by Revue