- Rapid Rundown
- Posts
- š± This unicorn startup faked 95% of its users
š± This unicorn startup faked 95% of its users
Plus: š The Most Advanced Wearable Device Ever, š„ Robots Watch, Robots Learn!
š¤ Investorās Nightmare: 95% of Reported Users Turn Out to Be Bots
Ironically, the social app IRL (In Real Life) had users that didnāt exist in real life.
An internal investigation found that 95% of the appās reported 20 million users were bots. After raising more than $200 million in venture capital, IRL is shutting down.
The company had its fair share of troubles, including layoffs and an SEC investigation.
IRL was supposed to be the next big thing for Gen Z event organizing, but after raising a SoftBank-led $170 million Series C round at a $1.17 billion valuation, things started to go south.
In the end, it seems that IRLās dreams of becoming an event-organizing alternative for Gen Z were just that - dreams.
š Humane Ai Pin: The Mysterious Gadget from Ex-Apple Employees Set to Disrupt the Tech Industry
Humane, a company started by former Apple employees, has announced its first gadget - the Humane Ai Pin.
Powered by an advanced Snapdragon platform in partnership with Qualcomm, the Ai Pin is set to launch later this year.
But what exactly does it do? Well, thatās still a mystery.
The company has been tight-lipped about the details, leaving us all to wonder about its capabilities. One thing we do know is that itās a āclothing-based wearable deviceā that uses AI and a range of sensors.
Itās a bold move away from screens and towards a new way of interacting with technology. We canāt wait to see what it can do!
š„ Unlocking Robot Learning: How YouTube Videos Are the Key
Robots have been trying to learn for decades, but itās not as easy as it sounds.
They need to do more than just follow programming - they need to adapt and learn. And it turns out, videos might be the key.
A team from CMU has developed a new system called VRB (Vision-Robotics Bridge) that lets robots learn from any video, not just ones that match their environment.
The system looks for clues like where the human touches the object and how they move it. Then it tries to copy them.
The best part is that the system can use any video, not just ones made for robots. The team uses videos from online databases that have thousands of hours of footage of humans doing everyday activities including the ones on YouTube.