Mad Mannequins, DIY Butler & The Sorting Robot 🤖

There was obviously a need for a secondary market for these mannequins, but the stores just didn’t care to find such resale markets. 

So Judi Henderson founded Mannequin Madness in her 50s. She started buying mannequins off craigslist and loaded an inventory of up to 500 mannequins within 6-months. The company first began renting and selling those mannequins. It then set up a photo studio for those who wanted to take photos with the dressed up mannies. 

In a few short years, Henderson set up a network across U.S. cities to recycle and upcycle mannequins, all of it at no-cost to the department stores and fashion brands trying to get rid of the mannies. The company averages around 300,000 lbs of mannequins per quarter. They’re on target this year to reach $1 million in sales.

Robots are being increasingly deployed in supply chain processes around the world to cope with the ongoing crisis, the rise in online shopping driven by the pandemic, and the lack of labour driven by the great resignation. 

In Greece, robots are sorting mail. 
55 robots work in the Athens sorting office. They weigh and scan the postal code on each package, then drop each parcel into the correct mail sack. They take a 15-min break every 4 hours to recharge. They process up to 168,000 parcels a day, increasing sorting speed by 250%.

Chinese online retailer JD.com uses hundreds of self-driving robots for delivery in several cities. 

Robots developed in Israel can pick up and move heavy boxes in warehouses even from the highest shelves.

Synthetic data looks or acts like it’s been collected from real people, when really it’s been created by artificial intelligence. 

It’s pretend-data. It’s fake-data. 

How It Works: 

Algorithms read real world data, flag patterns and trends, and then apply those patterns in real-world tasks such as targeted search results, targeted marketing, etc. 

Synthetic data hopes to break this targeting. In the new system, the algorithm will continue to map patterns and trends, which it will then use to create an entirely new set of data (pretend data) without any ties to real individuals.

Why Should You Care: 

For one, it’s a solution to the privacy problem.

Two, data collected from the real world typically reflects racism, sexism, and other forms of bias in society—and the algorithms learn and propagate those patterns on a large scale. There are numerous studies on how Google and Meta's biased algorithms influence micro and macro aspects of a person’s life such as his choices, job prospects, chances of getting into a good school or college. Basically, AI+data can now plan your whole life track without you realising it. Synthetic data could help create more “fair data algorithms” to deal with this bias.

According to a Gartner prediction, in the next two years, 60% of AI training data could be synthetic. 

Investment into synthetic-data startups is also flowing: 

• In October, Facebook acquired AI.Reverie. 
• In January, Mostly AI raised $25 million. 
• Synthetaic raised $13 million. 
• Israeli startup Datagen has raised $22 million and announced another $50 million funding round.

There are a bunch of other startups apart from these. All the startups have been around for only 4-5 years, so we’re really at the dawn of this new market. 

In 2021, the synthetic data market was worth $110 million, projected to reach $1.15 billion by 2027. Early adopters of the system include highly regulated industries such as health care, insurance, banking, and telecommunications.