Waymo patents soft self-driving cars. Esther Dyon + Jason Calacanis. Robot helps combat loneliness for ill children. http://cognitionx.com/news-briefing/.
Is OpenAI’s Dota 2-playing bot the new Alpha Go?
This past week, OpenAI’s bot went undefeated against the world’s greatest Dota 2 players. Dota is a complex game with hidden information. Agents must learn to plan, attack, trick, and deceive their opponents.
OpenAI argues that this win is a step towards building AI systems which accomplish well-defined goals in messy, complicated situations involving real humans.
Game-playing bots are not only of concern for gamers. As DeepMind writes, “Testing our agents in games that are not specifically designed for AI research, and where humans play well, is crucial to benchmark agent performance [in general]”.
Tabitha UntilTheBotsTakeOver Goldstaub
Considering implementing AI in your business? Satisfaction on the front lines is high: 62% of early adopters reported that outcomes from their AI initiatives exceeded their expectations. Drop us a line, we’d love to help you get there.
Karen Dolva, co-founder of the Norwegian startup No Isolation and her two co-founders – Marius Aabel and Matias Doyle – are all about utility. As their company name suggests, they are looking to end human isolation. It’s a massive undertaking, but they’ve started with a distinct and overlooked group: sick children.
Dolva realised just how important and neglected the issue of social solitude was when she met a woman who lost her teenage daughter to cancer. She and her partners researched the problem, speaking to children with a multitude of different health conditions and came up with an answer: a telepresence robot called AV1. Check it out on their website.
Waymo has just received a patent for an autonomous vehicle that becomes less rigid when its sensors tell it that it’s about to get into a collision. “The force of the vehicle’s impact is a primary factor in the amount of damage that is caused by the vehicle,” said the patent granted Aug. 8. “Accordingly, it is desirable to design a vehicle that can reduce the force of impact experienced during a collision.”
Under the newly patented technology, a car’s hood,
panels, and bumpers could be made rigid with the use of cables under tension. Reducing the tension would reduce the rigidity of the car in the place it was going to be hit, according to the patent.
Jason Calacanis interviews Esther Dyson, journalist, CognitionX advisor, and technology investor turned wellness enthusiast, who is on a mission to better health by rethinking and recreating the environment in which it can thrive. In this special episode of TWiST, Jason and Esther discuss the progression of angel investing over the years, the complexities around U.S. and Russia relations, predictions of AR, VR, & self-driving cars, real-world issues including diabetes and income volatility, and more.
This essay by Professor Ryan Calo, prepared in connection with UC Davis Law Review’s 50th anniversary symposium, explains why AI is suddenly on everyone’s mind and provides a roadmap to the major policy questions AI raises. The essay is designed to help policymakers, investors, technologists, scholars, and students understand the contemporary policy environment around AI at least well enough to initiative their own exploration.
Topics covered include:
• Justice and equity
• Use of force
• Safety and certification
• Privacy (including data parity); and
• Taxation and displacement of labor
Researchers developed an algorithm to reveal how much info you’re handing over when you click around, essentially picking out which parts of a recommendation engine are and aren’t safe for privacy-minded users.
The team in their paper fed their algorithm datasets from recommendation engines Movielens and Jester. Essentially, their math weighs the value users get by clicking against the personal info they’re disclosing, and points out when users would be handing over too much information for too little effect. While not everyone uses those particular services, they’ve definitely used a recommendation before if they’ve spun up a Pandora station or let Netflix suggest films.
On-demand food company Deliveroo has revealed that a new algorithm it calls ‘Frank’ has cut delivery times by 20% for its restaurants and riders. The ‘Frank’ algorithm is based on predictive technology that evaluates the most efficient way of distributing orders by location.
According to Deliveroo CTO Mike Hudack, machine learning is helping to predict the time it will take to prepare a meal, based on a specific type of order as well as on the time of day, all helping to streamline the experience from order to restaurant to customer. “We’re building an algorithm which can cope with millions of simultaneous orders,” said Hudack. “Our investment in the engineering team is testament to that ambition, and we are working on new, trailblazing technology that will make that ambition a reality.”
Robbie Allen writes about how to make yourself sound like you’re an expert in AI. Although it is written in a tongue-in-cheek manner, it’s a useful guide to stay on top of AI lingo and goings-on. Tips include:
Sporadically reference a paper on Arxiv as if everyone has read it
Propose your own “Net”
Trivialize existing “AI” companies
Disagree/agree with Elon Musk
On your way out, make a comment about automating Venture Capitalists