Artificial Intelligence/Machine Learning Roundup #16

1. [1805.01934] Learning to See in the Dark

Read More

2. [1805.03379] Opinion Fraud Detection via Neural Autoencoder Decision Forest

Read More

3. [1805.03545] Solving Sudoku with Ant Colony Optimisation

Read More

4. [1805.03616] A Reinforced Topic-Aware Convolutional Sequence-to-Sequence Model for Abstractive Text Summarization

Read More

5. New short film: The Internet of Things – what’s the big deal? Futurist Gerd Leonhard: conversation with Gerd (#3) – Futurist Gerd Leonhard

Read More

6. What are examples of AI coming up with unconventional solutions? – Quora

When I was working for Berkeley’s BEST lab on Tensegrity robots for NASA space missions, I was in charge of path-planning.

What you have to know is that this kind of robot moves by a deformation-based crawling/rolling technique (video). It can absorb shocks thanks to its springy structure, and would therefore have a thruster in its center allow it to hop. Read More

7. Google Stole Our Idea! – Little World: Inside John Done –

Yesterday, the Google Assistant team announced Google Duplex, a system to interact with businesses on your behalf over the telephone, using natural language. Here’s the video:

Here I am with my former co-founder Jeff doing a demo of John Done at last year’s Betaworks Voice Summit:

As best as we can surmise, here’s the Google Duplex roadmap:

Here’s the John Done roadmap circa July 2017:

As several friends have asked in the last 24 hours, “How does it feel when Google steals your idea?” OK, OK, I’ll tell you. Read More

8. How artificial intelligence can be applied to the real world

Artificial intelligence sounds like it belongs to the domain of science fiction, but companies have been using AI in very real and practical ways.

A new report from the McKinsey Global Institute titled “Notes from the AI Frontier” looks at how companies are currently using AI techniques to solve business problems along with how they can use them in the future.

“One of the things we’ve done is we’ve looked at something like over 500 use cases,” said James Manyika, director of the McKinsey Global Institute. Read More

9. Artificial intelligence: The advantages and disadvantages of AI

Australia and other countries around the world are seeing an increased use of artificial intelligence (AI).
RMIT Senior Lecturer Dr Andy Song believes the definition has been debated for quite a long time.

“There is no premise definition,” Dr Song said. Read More

10. Why good AI should be able to show its work

The first step to ethical artificial intelligence is teaching the computer to explain its decision making, something known in the field as explainable AI.

Why it matters: Right now many deep learning algorithms don’t make it clear how they arrived at their predictions or conclusions. That lack of visibility into the data, steps and calculations that went into an outcome makes it hard to root out bias or other algorithmic errors that could impact results like who gets a loan or how much a factory should produce. Read More

11. Kasparov: Self-driving won’t be perfect

When chess grandmaster Garry Kasparov considers self-driving cars, he thinks of elevators: for the first half century they were in use, almost no one would get on one unless an operator was aboard to push the buttons.

Why it matters: With a soothing ad campaign, the industry got people to stop sweating, and push their own buttons. So it is with autonomous driving, Kasparov told the WSJ Future of Everything Festival yesterday in New York: “No machine will reach 100% perfection, but it will be much safer. Read More

12. Are intelligent robots really intelligent?

With advances in technology robots are increasingly being used to do human jobs.

But are they actually better at doing them.
Will Gompertz goes to the exhibition exploring just that. Read More

13. Hey, Google, preheat the oven: More commands coming to Whirlpool appliances

It’s going to get a little bit easier to talk to your Whirlpool dishwasher, thanks to new enhancements to the Google Assistant voice-command features with Whirlpool’s connected appliance line.

Whirlpool already features integration with Google Assistant, but new features announced this week include the introduction of more concise functional Google Home commands and the expansion of remote appliance control to all regions where Whirlpool Connected appliances are available. ” The functional command upgrade applies to 25 connected appliances from the Whirlpool and Jenn-Air brands, which will carry official certifications as “Works with Google Assistant” appliances. Read More

14. Smart doorbells that call the police could put innocent people in danger

Smart doorbells — a gadget equipped with livestream and build-in speakers that allows you to answer your door when you’re not physically around — started as an innocent way to check in on your front porch when your Amazon package was being dropped off while you were at the office.

Amazon-owned smart doorbell company Ring just released an app called Neighbors that gives you the tools well beyond what is needed to check that your delivery made it to your porch. The standalone app allows you to get in touch directly with law enforcement — a local police force, or private security company — and notify them of anything fishy going on in your neighborhood, as Motherboard reports. Read More

15. DARPA’s Semi-Disposable Gremlin Drones Will Fly by 2019

We first reported on Gremlins back in 2015, as one of those “DARPA wants” projects that seems like it might be a bit far-fetched—in this case, DARPA wanted swarms of nearly disposable UAVs that could launch and be retrieved from flying aircraft carrier motherships in mid-air. Over the last few years, we’ve seen some progress towards disposable drones, but the tricky part was always going to be the mid-air retrieval. Each Gremlin will be designed to function for perhaps 20 missions, with minor refurbishment between missions. Read More

16. Mouth Sensor Can Measure the Salt in Every Potato Chip You Eat

If you’re one of the approximately 36 million adults in the United States with high blood pressure (also called hypertension), your doctor may have lectured you about the importance of reducing your sodium intake. The device, which they described in the journal PNAS, could send info to your phone, giving you instant data about whether you’re busting your low-sodium diet. Hong Yeo, an assistant professor of micro and nano engineering who led the research team, says it would also be possible to stick the sensor directly to the tongue or the roof of the mouth, or to laminate it onto a tooth. Read More

17. Cloud computing: Why this airline just went ‘all in’ on AWS

Video: Portrait of a modern multi-cloud data center.

Low-cost airline Ryanair plans to close the vast majority of its data centers over the next three years, as it moves its infrastructure to the cloud with Amazon Web Services (AWS). data center decision (free PDF)

Ryanair is standardizing on AWS services, including AWS databases, analytics, machine learning, and deep-learning services. Read More

18. This company audits algorithms to see how biased they are

Mathematician Cathy O’Neil is offering businesses a chance to test their algorithms for fairness.

Opening the black box: As artificial-intelligence systems get more advanced, the logic paths they follow can be difficult or even impossible to understand, creating a so-called “black box. Companies aren’t knocking down her door yet (she has only six clients). Read More

19. AI program gets really good at navigation by developing a brain-like GPS system

An AI program trained to navigate through a virtual maze has unexpectedly developed an architecture that resembles the neural “GPS system” found inside a brain.

The discovery comes from DeepMind, a UK company owned by Alphabet and dedicated to advancing general artificial intelligence. But this idea should be treated with some caution, since there is much we do not know about how the brain works, and since the functioning of artificial neural networks is also often hard to explain. Read More

20. Calls From Google’s “Duplex” System Should Include Initial Warning Announcements – Lauren Weinstein’s Blog

I’ve already received many reactions to my posting yesterday of “Google’s New ‘Duplex’ Voice System Creates Some Troublesome Questions” (https://lauren. com/2018/05/08/googles-new-duplex-voice-system-creates-some-troublesome-questions) — discussing Google’s new system capable of conducting telephone conversations that may be indistinguishable from human callers.

With no exceptions so far, the sense of these reactions has confirmed what I suspected — that people are just fine with talking to automated systems so long as they are aware of the fact that they are not talking to another person. Read More

21. Somewhere, a man-bun just died: Robot baristas are coming

Sorry human latté-artists, it’s time to make way for… robots.

From the team that brought you Beats by Dre comes… Cafe X.

It’s a new,$25k automated barista designed by the Ammunition Group that, according to CNBC, “slings” 120 cups of hot cawffee an hour. Read More

22. Google’s Duplex voice bot is clever, but I’m probably never going to use it

Google’s Duplex tech, which sees an AI mimic a human voice to make phone calls for you and do things like book an appointment at a hair salon, drew a massive ovation at the company’s I/O 2018 conference.

For the past couple of years, I have hardly had to make calls to book anything in my hometown of Bangalore, India; I rely on apps and sites to get event tickets, book flights and hotels, order food and groceries, and to get my home plumbing fixed.

Naturally, there are concerns about the ethics of using bots that can trick people into thinking they’re speaking to humans, and about the dangers of such technologies falling into the wrong hands (think automated spam calls on a grand scale). Read More

23. Can AI make the world a more inclusive place?

With a billion people worldwide living with some form of disability, there is tremendous scope for the development of assistive technologies – a market expected be worth over $26 Billion by 2024. AI will enable these devides to listen, see, reason and predict without a 24/7 dependence on the cloud (this physical interface between humans and machines is what Microsoft terms the “intelligent edge”) and the average smart home will generate around 50GB of data every single day. The program comprises grants, technology investments and expertise, with innovations integrated into Microsoft Cloud services. Read More

24. IoT botnet Hide and Seek evolves to survive device reboot


Security researchers have discovered the re-emergence of the Hide and Seek IoT botnet, which is the first of its type to be able to survive a device reboot.

Initially discovered in January this year by Bitdefender, the botnet has managed to infect around 90,000 devices worldwide. However, the Hide and Seek botnet is able to survive this, as long as the infection took place via Telnet. Read More

25. The Research Behind Google’s New Tools for Digital Well-Being

Google wants to help its users take back their time. ” Popularized by Tristan Harris, a former product manager at Google and cofounder of the Center for Humane Technology, TWS has snowballed into a movement, reaching acronymic levels of recognition in tech circles and becoming a major talking point among the industry’s biggest companies.

At first blush, Dashboard sounds a lot like Moment, a popular screen-time-tracking app designed by Kevin Holesh that’s been available for iOS since 2014. Read More

26. Facial Recognition Tech Is Creepy When It Works—And Creepier When It Doesn’t

For the last few years, police forces around China have invested heavily to build the world’s largest video surveillance and facial recognition system, incorporating more than 170 million cameras so far.

But the technology’s accuracy and reliability at this point is much more modest than advertised, and those imperfections make law enforcement’s use of it potentially sinister in a different way. ”

Those worried about how facial recognition surveillance may impact their personal privacy may view these flaws as a potential advantage; a fallible system might be easier to hide from. Read More

27. Google’s new conversational AI could eventually undermine our sense of identity

Read More

28. Graphical Representation of GANs Making New Molecules

Thursday, May 10, 2018, Baltimore, MD – Insilico Medicine, a Baltimore-based next-generation artificial intelligence company specializing in the application of deep learning for target identification, drug discovery and aging research announces the publication of a new research paper in Molecular Pharmaceutics journal titled “Adversarial Threshold Neural Computer for Molecular De Novo Design”. The described Adversarial Threshold Neural Computer (ATNC) model based on the combination of Generative Adversarial Networks (GANs) with Reinforcement Learning (RL) is intended for the design of novel small organic molecules with the desired set of pharmacological properties. I hope that part of this work integrated into our pipeline will help make the world a better and healthier place and help make perfect molecules for specific targets and multiple targets that will have a much higher chance of becoming great drugs”, said Evgeny Putin, the deep learning lead at Insilico Medicine. Read More

29. The Future Has Begun: Using Artificial Intelligence to Transform Government

Government agencies turn to artificial intelligence to buy equipment for warfighters, make American workplaces safer and much more.

By Peter Kamocsai, Associate Manager, Partnership for Public Service

Google held an event earlier this week to demonstrate an artificial intelligence-powered voice service by having it call a hair salon for an appointment.

Federal agencies were benefiting from AI long before the White House summit. Read More

30. What do we truly need? Emotional Intelligence or a more Technological Inclined Relationship?

What do we truly need.
With over Seven (7) Billion people on Earth, people are still LONELY. People are giving us what they think we want, not what we truly desire. Read More

31. Are Machines Capable of Beating Humans? – Tarun Kaushik –

Are Machines Capable of Beating Humans. Computers are made intelligent by using artificial neural network models. Due to machine learning algorithms, systems can perform analysis and give more accurate results. Read More

32. How the Google IO hairdresser call demo broke out the internet and yet we are still under…

Though if you have ever tried to have a conversation with your voice assistant after 2 to 3 sentences things can start to go wrong and our voice assistant gets confused. We are just in the early days of these experiences and have to be patience, we all remember the Nokia 3210 and can see how far things have come since those days. The Internet then started to freak out as it does when these announcments are made, though this demo was impressive it was firstly a demo and secondly when you really think about it, the problem is a fairly easy task for the assistant to perform. Read More

33. Artificial Intelligence In Search of Protection — Part IV

If you are an AI entrepreneur that wants to try to patent something, it is certainly useful to start from mapping the landscape of existing patents to see whether something similar has already been filed and whether an opportunity materializes for your innovation. In particular, in the last five years, the issue is becoming more tricky and the US IP law started drawing some clear lines (the most famous case up to date is Alice Corp. , neural networks, training sets, AI-generated code, learning algorithm, etc. Read More

34. The Great AI Debate – The Impact of AI on News and Media

Read More

35. Opinion: Google’s use of AI to mimic humans is unethical and bad UX

Google CEO Sundar Pichai took a giant leap in the wrong direction yesterday. At his company’s I/O developer conference, Pichai wowed the crowd by using a virtual assistant to fool people, making unsuspecting humans the target of laughter and derision and raising serious ethical questions about future uses for AI.

Pichai played for the audience a conversation between Google Assistant and a hair salon assistant:

The virtual assistant’s utterance of “Mm-hmm” generated peals of laughter from those eavesdropping on the equivalent of a prank call. Read More

36. Do These Three Things and You Won’t Lose Your Job to A.I.

Your job will be automated.

It’s not just hype; McKinsey estimates that 45% of activities in the American job market can be automated. Read More

37. Fei-Fei Li at Google I/O: Humans Overestimate AI in the Short-Term, Underestimate Its Long-Term…

Fei-Fei Li at Google I/O: Humans Overestimate AI in the Short-Term, Underestimate Its Long-Term Potential

Google Cloud Chief Scientist Fei-Fei Li is one of the most popular and influential AI figures today.

Li was joined by Google Principal Scientist Greg Corrado and Google Cloud CEO Diane Greene on the “Building the future of artificial intelligence for everyone” panel. The trio shared their views on the future of AI from a scientist’s perspective. Read More

38. One problem to explain why AI works –

A framework for understanding AI and keeping up with it all. Why will an expensive and inscrutable machine create the knowledge I need to solve my problem.

Part 1: Induction as the prevailing theory

Many overlook the question because it’s obvious how knowledge is created: We learn from observation. Read More

39. The Future Will Be Synthesized – Matt Hartman –

On the Authenticity Awareness Gap & Near-Term Future of Synthetic Media

There is a new media type emerging at the intersection of computer generated images, video, and voice. Today the applications are entertaining, but we are also starting to see the first examples of how this new technology can be used to create forged videos and audio clips — showing people saying things they didn’t actually say, and even doing things they didn’t actually do.

A (Very) Brief History of Photo Manipulation

Synthetic media is not new. Read More

40. Five Lessons from the O’Reilly AI Conference –

Three of us from Upside (myself, Billy Laing, and Rav Pasricha) went to NYC for the 3rd Annual O’Reilly AI Conference at the NYC Hilton Midtown for two days. Between the three of us, we covered about 20 different sessions in addition to attending the usual networking events and vendor expo. This is Computer Science’s main event. Read More

41. This is Why Algorithms May Replace Humans In Workforce

Read More

42. Five ways artificial intelligence will shape the future of universities

Within a few short years, universities may well have changed beyond all recognition. Here are five ways that AI will help to change and shape the future of universities and higher education for the better. Personalised learning

Universities are already using AI algorithms to personalise learning and deliver content that is suited to the students needs and pace of learning – and this is only likely to continue. Read More

43. Alexa and Siri Can Hear This Hidden Command. You Can’t.

— Many people have grown accustomed to talking to their smart devices, asking them to read a text, play a song or set an alarm.

Over the past two years, researchers in China and the United States have begun demonstrating that they can send hidden commands that are undetectable to the human ear to Apple’s Siri, Amazon’s Alexa and Google’s Assistant. In the wrong hands, the technology could be used to unlock doors, wire money or buy stuff online — simply with music playing over the radio. Read More

44. ICOAI 2018 : 2018 5th International Conference on Artificial Intelligence (ICOAI 2018)

Read More

45. Clouds, catapults and life after the end of Moore’s Law with Dr. Doug Burger

Some of the world’s leading architects are people that you’ve probably never heard of, and they’ve designed and built some of the world’s most amazing structures that you’ve probably never seen. Doug Burger, Distinguished Engineer at Microsoft Research NExT. And when you see a design that’s done well it’s beautiful. Read More

46. Google I/O Day 2 Highlights: How AI, machine learning can save more lives by making healthcare predictive

Somewhere in our life we have, unfortunately, heard a doctor express with utmost despair that they could have done better if the patient could have been brought to medical attention a little early-on. Google

thinks it can solve this problem for mankind.

Last year at Google I/O, Sundar Pichai

demonstrated the advanced computing powers of Google named Tensor Processing Units (TPU) which now has been working with eye hospitals like Arvind and Shankara in India

to help doctors use Deep Learning, a machine learning

module to screen Diabetic Retinopathy in a better way. Read More

47. What Pixar Can Teach Us About AI & Machine Learning

Read More

48. How To Talk To Plants Using Machine Learning And Gesture Recognition

Read More

49. Intel’s A.I. Architect Singer Lays Out the Vision for Deep Learning

Read More

50. Google Upgrades Its Speech-to-Text Service with Tailored Deep-Learning Models

A month after Google announced breakthroughs in Text-to-Speech generation technologies stemming from the Magenta project, the company followed through with a major upgrade of its Speech-to-Text API cloud service. The updated service leverages deep-learning models for speech transcription that are tailored to specific use-cases: short voice commands, phone calls and video, with a default model in all other contexts.

The specialized models are adapted to the characteristics of the audio media in terms of sampling, resulting in bandwidth and signal duration. Read More

Updated with 2 more links:

Sex-crime, trafficking and Machine Learning: Avengers at last

Robot designed to keep the elderly company