Posted in AI advances

Samsung uses AI to transform photos into talking head videos | ZDNET

Still images come to life using a AI model developed by Samsung at its Moscow research lab, ZDNET reports. Speech-like sequences can be generated from reference images as few as a single frame, such as a rare photo. Researchers claim the results can be…

Posted in AI advances

Teaching machines to write better adverts than humans | BBC

A machine that generates 20 to 25 ads per second — no extra time or charge for the puns — is one of the advances in AI systems that ‘create’ narratives. BBC News reports some of the call-to-action phrases can be more effective than…

Posted in AI advances

A poetry-writing AI has just been unveiled. It’s … pretty good. | VOX

VOX tests the latest version of GPT-2, the controversial text-generating algorithm, and now you can too. A Toronto machine language engineer has put together a simple website so anyone can give try the GPT-2 ‘writing’ algorithm using its most recently updated model. GPT-2 caused…

Posted in AI advances

Drowning in Research Reading? AI Could Help | INSIDE HIGHER ED

‘(It) could potentially be used by journalists to help communicate complex research to the public, though the authors say they aren’t going to be putting journalists out of a job any time soon.‘ A new algorithm for summarizing text is reported by INSIDE HIGHER…

Posted in AI advances

A new way to build tiny neural networks could create powerful AI on your phone | MIT TECHNOLOGY REVIEW

MIT Technology Review reports on a research discovery that suggests some AI capabilities can be achieved with simpler processing and in turn need less computing power. The findings come from looking into the workings of neural networks, structures of nodes, called neurons for their…

Posted in AI advances

OpenAI wants to make ultrapowerful AI, but not in a bad way | WIRED

WIRED reports on the business evolution of a Silicon Valley company fully committed to AI that betters or exceeds human performance. We’ve included elsewhere stories about OpenAI projects, in New AI fake text generator may be too dangerous to release, say creators | THE…

Posted in AI advances

AI model adds music to its repertoire | jAI

An AI technology that generates believable text now produces tunes, too. Both come from OpenAI, a California company dedicated to developing artificial general intelligence. Their newest model is called MuseNet. OpenAI says MuseNet can generate an original composition up to four minutes long, orchestrated…

Posted in AI advances

Brain signals translated into speech using artificial intelligence | NATURE

You could call it ‘speaking your mind’ — your words detected from your brainwaves, then decoded with an AI system and artificially produced, a machine saying what you wanted to say milliseconds earlier. Years from now, the system could provide an alternative to people…

Posted in AI advances

First AI-generated textbook published by Springer | jAI

A 247-page publication about lithium-ion batteries is the first textbook authored by a computer, setting a new milestone in machine-generated publishing. Springer, a leading publisher, collaborated with researchers at the Applied Computational Linguistics (ACoLi) lab at Goethe University in Frankfurt to produce the book….

Posted in AI advances

Amazon Alexa now offers long-form news coverage in addition to Flash Briefings | TECHCRUNCH

46% of smart speaker owners ask for news and segment leader Amazon has begun offering US users enriched news briefings, according to a report in TECH CRUNCH. Plans for other countries are not mentioned. Amazon says the new service combines news from Bloomberg, CNBC,…

Posted in AI advances Panels/presentations

iFlytek | AI-DAY 2019

Runs 27:35 Presentation by iFlytek, named by MIT in 2017 as the most innovative company in China and number six worldwide. Presenters: Zhang Wuxu, Introduction Xie Fei, General Manager of Fintech and Smart Service Business Unit Wang Wei, VP of Consumer Business Group March…

Posted in AI advances

The out-there AI ideas designed to keep the US ahead of China | MIT TECHNOLOGY REVIEW

MIT TECHNOLOGY REVIEW reports new details of a $2 billion plan for AI research by DARPA, the U.S. Defense Advanced Research Projects Agency. The spending is to advance the next generation of AI technology. The story says project areas include: teaching AI models how…

Posted in AI advances

Why AI is a threat to democracy—and what we can do to stop it | MIT TECHNOLOGY REVIEW

‘The problem is our technology has become more and more sophisticated, but our thinking on what is free speech and what does a free market economy look like has not become as sophisticated.’ Futurist Amy Webb MIT TECHNOLOGY REVIEW interviews futurist Amy Webb about…

Posted in AI advances Op-Eds

Jill Lepore: Are Robots Competing for Your Job? | February 2019

‘…in American history, panic about technological change is almost always tangled up with panic about immigration’ IN THE NEW YORKER Jill Lepore argues that advances in technology and having a job are two sides of the same coin. Lepore is a history professor at…

Posted in AI advances

A philosopher argues that an AI can’t be an artist | MIT TECHNOLOGY REVIEW

 ‘…the trumpet is not, itself, creative. Artificial-intelligence algorithms are more like musical instruments than they are like people..’  In MIT TECHNOLOGY REVIEW Harvard philosopher Sean Dorrance Kelly says creativity is about more than words, notes, or brush strokes and cannot be replaced by machines….

Posted in AI advances

As artificial intelligence changes the world, it changes our language too | LSE BUSINESS REVIEW

‘Usually the new words flow from necessity and are created organically. But with AI, I am going to jump the gun a bit and suggest some new words that we are going to need to adjust to this new technology.’ THE LONDON SCHOOL OF…

Posted in AI advances Op-Eds

Yael Eisenstat: The Real Reason Tech Struggles With Algorithmic Bias | Feb 2019

IN WIRED MAGAZINE former Facebook elections integrity head Yael Eisenstat observes the ‘move fast’ culture in tech means there is insufficient training for the people who are responsible for preventing algorithmic bias. ‘I did not know anyone who intentionally wanted to incorporate bias into…

Posted in AI advances

GANS: Faces from data

AI SCIENTIST IAN GOODFELLOW QUIETLY posted these five images to Twitter shortly into the new year, without any other statement except the links to papers describing each improvement. Keep in mind: none of these ‘people’ exist, except as rendered by a machine using a new…

Posted in AI advances Op-Eds

Steven Strogatz: One Giant Step for a Chess Playing Machine | Dec 2018

‘AlphaZero seemed to express insight. It played like no computer ever has, intuitively and beautifully, with a romantic, attacking style.’ IN THE NEW YORK TIMES mathematician Steven Strogatz says the threshold crossed by the game-playing AI, AlphaZero, displayed ‘a breed of intellect humans have…

Posted in AI advances Reports/white papers

Artificial Intelligence Index | 2018

AI research continues to be predominantly male-driven according to this annual index of key indicators compiled in the AI research community. REPRESENTATIVE FINDINGS Demographics of AI professors: 80% male (sample of seven leading faculties worldwide) Applicant pool for AI jobs in USA: 71% male…

Posted in AI advances

2018: 18 highlights in computer science

THE 18 BIGGEST DEVELOPMENTS IN COMPUTER SCIENCE FOR 2018 have been named by MIT’s Computer Science and AI Lab (CSAIL). They treat each equally. The list includes both new applications (e.g., cybernetic plants) and new outcomes (e.g. detection of metastatic cancer). Practices by Cambridge…

Posted in AI advances Quotes

Walsh on AI threats

‘I don’t worry about smart AI. I worry about stupid AI.’ – Toby Walsh

Posted in AI advances

AI has a culturally biased world view that Google has a plan to change | THE DOWNLOAD

A standard image-recognition system, trained on open-source data sets, can recognize a bride in a white dress, reflecting the classic Western tradition. But it will fail to recognize a bride in a sari from an Indian ceremony. THE DOWNLOAD, a publication of MIT, reports…

Posted in AI advances

The first AI-scripted commercial is here – and it’s surprisingly good | WEF

‘Watson was given 15 years’ worth of award-winning commercials, a trove of data showing when consumers tended to connect most emotionally with advertisements, and further data from an experiment conducted by the University of New South Wales’ MindX, which analyzed how highly intuitive people…

Posted in AI advances

China reveals world’s first AI anchor | SCMP

Runs :59 Text on video report by the South China Morning Post — the most comprehensive of the summaries prepared when Xinhua News Agency announced their first AI anchor. November 2018 

Posted in AI advances Book reviews

Walsh 2.0

AI expert Toby Walsh presents a sweeping inventory of changes ahead. The title is grounded in Walsh’s home turf of computer science. 2062 is the year by which there is a 50% probability that computers will match or exceed human abilities in most areas….

Posted in AI advances Op-Eds

Henry Kissinger: How the enlightenment ends | June 2018

‘By treating a mathematical process as if it were a thought process, and either trying to mimic that process ourselves or merely accepting the results, we are in danger of losing the capacity that has been the essence of human cognition.’ IN THE ATLANTIC…

Posted in AI advances

A DARPA perspective on artificial intelligence | DARPA

Runs 16:01 DARPA Director of Information Innovation John Launchbury gives a even-handed explanation of AI capabilities: what it can do, what it can’t do, and where it is headed. He describes AI in three waves: 1st wave – Handcrafted knowledge (past and present applications)…

Posted in AI advances Book reviews

Baseline

AI professor Toby Walsh is convinced big changes are ahead. Walsh is known as a ‘rock star’ in the Australian digital community and internationally as an advocate for AI that improves human life. He also is a high profile activist against intelligent weaponry. He grounds his…

Posted in AI advances Panels/presentations

Asimov Debate 2018

Runs 122m PANEL moderated by Neil deGrasse Tyson, astrophysicist and head of the Hayden Planetarium. PANELISTS: John Giannandrea – Senior Vice President of Engineering at Google, where he leads the Search and Google A.I. teams.  Helen Greiner – co-founder of the iRobot Corporation, maker of the Roomba Vacuuming Robot….

Posted in AI advances Panels/presentations

Technology, Media & Democracy | MACHINES + MEDIA 2018

Runs 37:57 PANEL moderated by John Micklethwait, Editor-in-Chief, Bloomberg PANELISTS Jessica Rosenworcel, Federal Communications Commission Clay Shirky, New York University February 2018 | New York

Posted in AI advances Panels/presentations

Alexa State of the Science | Amazon

Runs 49:39 TALK by Alexa VP and Head Scientist Rohit Prasad, explaining how the voice assistant presents ‘invisible AI’ to an Amazon audience. Although recorded in November 2017, it’s useful for the foundation concepts he describes. NOTE: Rohit’s presentation runs approx. 32 minutes, followed…

Posted in AI advances

GANS: How machines convert data to images

Runs 21:20 Rob Miles is a Ph.D. student in computer science and a YouTuber about AI. In this video he walks through the processes involved in Generative Adversarial Networks (GANS). A lot of AI activity involves classifying, for example letters, images, or sounds that…

Posted in AI advances Book reviews

The technological stage

Life 3.0 opens with sentient AIs controlling news and media outlets. Tegmark is a professor of physics at M.I.T., author of more than 200 scientific papers, and founder of the Future of Life Institute, an international group that advocates beneficial uses of artificial intelligence….

Posted in Academic papers AI advances

When Will AI Exceed Human Performance? Evidence from AI Experts | Grace et al

When Katja Grace and co-authors first released this paper in June 2017, it provoked worldwide headlines. Their findings showed AI experts believe there is a 50% probability AI will ‘outperform humans in all tasks’ by 2062. The paper presents a timeline of dates by…