What is it?
Artificial Intelligence (AI) is about simulating reasoning, developing knowledge, and allowing computers to set and achieve goals through a wide range of techniques such as machine learning, deep learning, probabilistic inference, neural network simulation, pattern analysis, decision trees, and random forests. And while today’s leading AI approaches represent supervised learning with narrow focus, it appears the goal is to push, for general purpose, intelligence that can be self-taught and self-learning.
Here is a snapshot of the machine intelligence landscape from Bloomberg Beta's Shivon Zilis:
The latest developments on AI can be found in these blogs by Peter Xing (originally appeared on Deloitte Digital):
As set out in our last instalment in January, this year was going to see a further acceleration in developments for AI. Here are some of the highlights that have come to pass:
1. NVIDIA goes all in on deep learning at GTC 2016
As visited in last April’s blog, Nvidia have made it clear on their strategy to invest heavily in AI. At GTC 2016, Nvidia’s CEO, Jen-Hsun Huang unveiled the DGX-1 - purportedly the world's first deep learning supercomputer which reaches 170 teraflops (a trillion floating point operations per second) with a pricetag of $129,000. With 8 tightly-coupled state-of-the-art GPUs, it is nearly 10 times faster at supervised learning than Nvidia’s flagship unit a year ago.
Nvidia also unveiled DAVENET, a deep neural network that has been trained to know how to drive. Traditionally, autonomous vehicles, such as the ones used in the DARPA challenge, have relied on manually-coded algorithms to follow a desired route, and provide vehicle control. Nvidia (along with many other current vehicle research teams) has been experimenting with using deep learning neural networks instead. According to Huang (and illustrated with a demo video), after only 3,000 miles of supervised driving, its car was able to navigate on freeways, country roads, gravel driveways, and in the rain.
Huang then went on to announce Roborace - a fully autonomous racing league. They will race on the same courses as Formula E, except without drivers. The teams’ innovation and differentiation will be in the software they develop for the race. Roborace founder Dennis Sverdlov said “it’s not possible to get competitive advantage based on how much money you put in hardware. Our heroes are not the drivers. Our heroes are engineers.”
2. Salesforce acquires MetaMind
MetaMind, a Palo Alto-based AI startup founded in July 2014, is being acquired by Salesforce. MetaMind’s CEO, Richard Socher is a Stanford PhD who studied machine learning, deep learning, natural language processing and computer vision under Baidu’s Andrew Ng. Salesforce plans to use MetaMind’s technology to “further automate and personalise customer support, marketing automation, and many other business processes. [MetaMind will] extend Salesforce’s data science capabilities by embedding deep learning within the Salesforce platform.”
As a standalone company, MetaMind’s general-purpose platform was designed to predict outcomes for language, vision and database tasks. As of the middle of last year, its technology could reportedly answer everything from specific queries about snippets of text to the sentiment of that text.
MetaMind had raised $8 million from investors, according to CrunchBase. Its backers include Khosla Ventures and (notably) Salesforce CEO Marc Benioff.
3. Google Deepmind’s AlphaGo defeats world champion
After beating the European champion, shocking experts who predicted it wouldn’t happen for another decade, Google DeepMind’s deep learning AI AlphaGo went on to defeat the world champion just a month later, winning 4-1. "I've never been congratulated so much just because I won one game” was the reaction from AlphaGo’s opponent, Lee Sedol. AlphaGo initially was programmed by learning thousands of human games, but later on it used deep reinforcement learning and improved by playing itself. Some of it's moves are therefore not human at all and surprise even seasoned players. There are more possible moves in a game of Go than there are atoms in the entire universe
With all this coverage, a Chinese team is now challenging AlphaGo for a match towards the end of this year. Meanwhile, Deepmind have launched its Health division, following the pursuits of IBM Watson in tackling the healthcare industry.
4. A novel written by AI passes the first round in a Japanese literary competition
“The Day A Computer Writes A Novel” was a short story authored by an AI that has made it through to the latter stages of a literary competition in Japan. The short-form novel was written with the help of a team of researchers from the Future University Hakodate in Japan. Human beings selected certain words and phrases to be used, and set up an overall framework for the story, before letting the software come up with the text itself. Of 1,450 or so novels accepted this year, 11 were written with the involvement of AI programs, the Japan News reports. Judges aren't told in advance which submissions are written by actual people and which have robot authors behind them.
Automated software is already responsible for writing certain financial and sports reports where the key facts can be arranged in a straightforward template. Political speeches are another target for up-and-coming robot writers, as they tend to follow a familiar pattern, with repeated phrases and topics. As is usually the case, the database the AI has to work with is crucial – as long as there's enough data to draw upon (4,000 speeches were used for the latest research), then today's AI software is clever enough to produce its own variations on a theme.
5. Developments for teen AI chatbot, Tay.AI a social commentary on humanity?
"Back to the drawing board." That's how Chief Executive Satya Nadella addressed Microsoft's recent gaffes with Tay, its AI teenage chatbot. Microsoft took the chatbot offline shortly after it spewed racist and sexist remarks on Twitter.
Microsoft has joined tech giants like Facebook, Google, Apple and IBM who are trying to create software that can learn from what we do and, by extension, help us in our daily lives. Facebook, for example, is teaching its AI how to recognise shapes in a photo so that it can tell blind users what's on the screen. However, learning from this experience, it appears there are many folks out there who can pose a mischief to these innocent learning chatbots.
6. Engineer builds robot that looks like Scarlett Johansson
Engineer Ricky Ma says he spent 18 months crafting his “Mark 1” robot. At a cost of $50,000, the robot is made from 3D-printed parts and silicone skin, and can display animated facial expressions such as smiling and winking. Mark 1 can also move her arms and legs and responds to Ma when he compliments her. Ma claims that his programming skills are self-taught. Although he encountered many obstacles while building Mark 1, he says he has no regrets, especially with the end result. Without admitting it, the robot has an uncanny resemblance to actor, Scarlett Johansson. Check out the video of Ma showing off his bot
If you’re curious to know when, if ever, AI can achieve human level consciousness, here’s a list of expert views on the matter.
In this New Year edition of Sightings for the Edge: Artificial Intelligence, we explore the round-ups of AI developments in 2015 and what to expect over the coming year.
1. AI in 2016 “Like 2015 on Steroids”
2015 was a pretty big year for AI. Shivon Zilis had to update her Machine Intelligence Landscape to 2.0 just for the new innovations. This is a must see:
The consensus is that 2016 is going to be even bigger. TechRepublic checked in with AI experts Andrew Moore, Kathleen Richardson, and Roman Yampolskiy, for their take on what we've seen in AI this year and what's coming in 2016.
From deep learning, to AI replacing workers, to IoT and breakthroughs in emotional understanding, the AI experts provide their view on what’s still to come.
2. Complete List of Useful AI Tools
Tired of hearing about how another AI product will change the world? Get past the vapourware and check out this list of truly useful AI tools you can use today from Bob Gurley at Cognitio Corp . From AI on your phone, cloud, personal, business, developer, healthcare, robots, space, marketing and customer service, Bob makes a solid attempt to capture all the AI you can use right now in his publication for CTOvision.com.
We’re putting together a detailed list of all AI startups, mature companies and products out there, sourcing originally from the Machine Intelligence Landscape from Shivon Zilis, and updated through AngelList, Crunchbase, Mattermark and the latest AI news and websites. See our work in progress on this Google Sheet.
Feel free to add to this list and let us know what you think about the usefulness of some of the AI products from this list.
3. Investing in AI
With AI development accelerating every day, the question on everyone’s mind is how do we make money on this stuff? Nathan Benaich at Playfair Capital contributes this piece to TechCrunch to provide an overview of what to consider when investing in AI.
4. Deep Learning Startups to Follow in 2016
VentureBeat has come up with a list of deep learning startups to follow in 2016:
Deep Instinct appliesdeep learning to get big in the antivirus software market.
Lunit helps radiologists understand medical images. U.S. startup Enlitic also does that, but Lunit is different by virtue of its Data-driven Imaging Biomarker (DIB) algorithm and its focus on chest X-rays.
Nnaisense is building a general purpose neural network-based artificial intelligence.
TeraDeep has built technology for image recognition, offering software and APIs that developers can incorporate into their applications.
Vuno has been busy forming partnerships with healthcare organisations to help doctors diagnose diseases with technology that makes inferences about medical data.
We’ll keep an eye on these and evaluate at the end of the year.
5. Top AI Breakthroughs of 2015
So what is fundamentally driving the plethora of AI startups appearing in the market? The Future of Life Institute explores the top AI breakthroughs of 2015 and provide on where we are heading if more breakthroughs become commonplace.
The Future of Life Institute is a volunteer-run research and outreach organization in the Boston area that works to mitigate existential risks facing humanity, particularly existential risk from advanced AI. Its founders include MIT cosmologist Max Tegmark and Skype co-founder Jaan Tallinn, and its board of advisors includes cosmologist Stephen Hawking and entrepreneur Elon Musk.
In this installment, we further explore the disruption of AI to enterprise functions and industries around the world.
1. Legal Services
Law firm bosses envision IBM Watson-type computers replacing young lawyers.
“The study, which included responses from high-ranking lawyers at 320 firms with at least 50 lawyers on staff, found that 35 per cent of the top brass at responding law firms envision replacing first-year associates with some type of AI in the coming decade. Less than 25 per cent of respondents gave the same answer in a similar survey in 2011.”
Facebook’s AI can caption photos for the blind and read it out to them, thanks to image recognition and text-to-speech software.
“As a blind user, going from essentially zero percent satisfaction from a photo to somewhere in the neighborhood of half ... is a huge jump.”
3. Search Engines
Not surprisingly, Google is turning its lucrative web search over to AI. RankBrain now handles a “very high fraction” of all Google search results.
“If RankBrain sees a word or phrase it isn’t familiar with, the machine can make a guess as to what words or phrases might have a similar meaning and filter the result accordingly, making it more effective at handling never-before-seen search queries.”
An antivirus that mimics the brain could catch more new malware using deep learning.
“Essentially it can tell if a file is similar enough to an existing piece of malware to make it suspicious. Existing antivirus software may be fooled if the particular string of code it is using for detection has been altered.”
5. Weather forecasting
IBM is about to become the best weather forecaster of all time after confirming that it will acquire the digital assets of The Weather Channel.
“IBM offers a cloud service that allows developers to upload and analyze data from various “Internet of Things” devices such as environmental sensors. Some of that data could find its way into the Weather Company’s models, if the owners of that data allow it.”
For more AI developments, follow me on twitter @peterxing.
In this installment, we further explore the emergence of AI in disrupting enterprise functions and industries around the world.
1. Medical and diagnostics
IBM is planning to acquire Merge Healthcare, a company that specialises in medical imaging software. “The goal is to bring enhanced imaging capabilities to the Watson Health portfolio, essentially giving the supercomputer the ability to see.”
Enlitic Inc., a San Francisco-based startup with $5 million in angel and seed funding, claims that its software identified malignant lung tumors in X-rays 50% more accurately than a panel of four radiologists.
Meanwhile, Baidu is working on a project called AskADoctor, and their “long-term goal is to build a medical robot” arising from the frustrations with China’s overburdened healthcare system.
2. Virtual Assistants
Slack aims for its Slackbot to be the “virtual assistant to everyone on a team simultaneously, while also having access to their organisation’s institutional knowledge: who is working on which projects, where they’re stuck, who is on vacation and when they’re getting back.” CEO Stewart Butterfield says that Slackbot could boost a company’s productivity by 20-30%”.
x.ai have developed a virtual personal assistant called Amy Ingram to schedule your meetings. Once enabled, all it takes is for your to cc Amy in your conversations.
“...Despite not being human, Amy was very realistic. Even when meetings needed to be rescheduled (from my end, or theirs,) Amy seemed to understand, and the next email I’d see from her would be an updated calendar invite.” - Lara O’Reily, Business Insider
Google have cut voice transcription error rates down from 16% to 8% by switching from an old machine learning technique to deep learning - specifically, Long Short-Term Memory Recurrent Neural Networks.
Tesla’s autopilot is about to steer the Model S.
“It's much like autopilot in a plane. You turn it on in a plane, but there's still some expectation the pilot will pay attention to what the plane is doing and won't go to sleep or disappear from the cockpit.” - Elon Musk.
Uber - At a panel during a Top 10 Tech Trends dinner in San Jose, Steve Jurvetson, an early Tesla investor and current company board member, stated that Uber’s CEO told him that if Tesla is able to build a fully-functioning autonomous vehicle by 2020, Uber would want to buy all 500,000 of ones estimated to be produced.
Apple - Documents confirm that Apple is building a driverless car.
However, Google’s driverless cars are probably miles ahead.
“And, unlike human drivers who must rely on their own experience for learning, Google’s cars will learn from every Google cars’ experience. That means that the more cars Google puts on the road compared to its competitors, the greater its learning advantage.”
4. Engineering and Design
A study by the University of Cambridge examined the “evolution” of robots, where a “mother” robot created increasingly successful offspring based on natural selection.
Here’s a video of the teaser.
Henn-na in Japan say they have developed the world’s first hotel run by robots
“The hotel’s owner, Hideo Sawada, says he wants to make this the most efficient hotel in the world by reducing manpower and having 90% of staff be robotic.”
The initial reviews, however, are a little disappointing.
“At 3pm the velociraptor jerks to life and says, in an American accent, ‘Welcome to the Henn-na hotel. If you want to check in, press one.’ I start tapping the screen but the man in black appears again and asks for my passport, leaving the robot to fall into a state of inertia.”
1. Gov2020: The Future of Artificial Intelligence
“The centuries’ long quest to develop machines and software with human-like intelligence inches closer to reality. Scientists develop intelligent machines that can simulate reasoning, develop knowledge, and allow computers to set and achieve goals, moving closer to mimicking the human thought process. These intelligent systems improve accuracy of predictions, accelerate problem solving and automate administrative tasks bringing in an era of automation.” –Deloitte Gov2020.
Driver Categories > Technology: Exponential >Artificial Intelligence
Gov2020 helps leaders from across the private and public sectors make sense of the rapidly changing demographic, societal, economic, and technological trends shaping the future. Gov2020 isn’t a crystal ball but it does pull together the best of Deloitte research and expertise from across the globe to start a discussion on what is probable, and even more importantly, what is possible.
2. TED Talks powered by Watson
“Imagine being able to ask a panel of TED speakers: Will having more money make me happy? Will new innovations give me a longer life? A new technology from IBM Watson is set to help people explore the ideas inside TED Talks videos - by asking the questions that matter to them, in natural language.” –watson.ted.com
Search through TED’s 1,900 talks, powered by IBM Watson. Read the blog, see the demo in action or try the alpha version yourself. A useful resource for the winner of the AI XPRIZE presented by TED!
3. Wibbitz | Automated Insights | Narrative Science
Wibbitz is a text-to-video production startup and it just raised $8M to replace your publication’s video team with robots!
Check out the real-time examples of Wibbitz-created videos here, using the power of text-to-video AI.
With the advent of robot journalists, the addition of engaging and relevant videos to automated news publications really makes this blogger worry…
And why stop at journalism?
Automated Insights writes insightful, personalised reports from your data. It's like an expert talking with each user in plain English. Similarly, Narrative Science writes stories based on your data to accelerate decision-making, enhance customer engagement and improve employee productivity.
Check out Narrative Science’s report on the State of AI & Big Data in the Enterprise, in which they’ve found that 80% of business and tech leaders say AI creates jobs and increases productivity. The leaders surveyed were humans with jobs, nevertheless!
Here’s 3 white collar jobs that AI and robotics are already mastering. Indeed, we need new jobs as the machines do more of our work.
4. Humans vs AI in Texas Hold’em
First, they beat us at chess. Then, Jeopardy. And now…Texas Hold’em?
Thankfully, the humans won… this time!
5. Deep Learning Machine Beats Humans in IQ Test
“Computers have never been good at answering the type of verbal reasoning questions found in IQ tests. Now a deep learning machine unveiled in China is changing that” –MIT Technology Review
“To our surprise, the average performance of human beings is a little lower than that of our proposed method” - Huazheng Wang at the University of Science and Technology of China and Bin Gao at Microsoft Research in Beijing.
6. AI Scientists
A new scientific theory has been invented by a computer without human help for the first time.
“Computer scientists from the University of Maryland programmed a computer to randomly predict how a worm's genes formed a regulatory network capable of regeneration, before evaluating these predictions through simulation. The computer invented an accurate model of the inner-workings of a flatworm. After three days of continuously predicting, simulating and evaluating, the computer was able to come up with a core genetic network that explained how the worm's regeneration took place”.
7. Cognitive technology - The rise of “bionic brains”
“Augmentation won’t save every job from automation, but it’s perhaps the best hope we have. Humans aren’t going to improve their brains at the rate computers will, and we’re not going to turn the bionic brains off. So we need to devote a good chunk of our brainpower and creativity to how we can work in partnership with the smart machines that we have created” – Tom Davenport for Deloitte University Press
8. Google’s Ray Kurzweil says humans will have 'hybrid' cloud-powered brains by 2030
Within 15 years, humans will be implanted with nanobots that will connect their brains to the internet, allowing for vastly accelerated cognition. Ten years after that, most of our thinking “will be done online,” according to futurologist Ray Kurzweil.
9. This AI Pioneer has a few concerns
An interview with the British-American computer scientist, Stuart Russell who drafted and became the first signatory of an open letter calling for researchers to look beyond the goal of merely making AI more powerful.
It begs the question of what does the rapid advance of AI mean for humanity?
10. Brain Implant Lets Tetraplegic Man Drink Beer with a Robot Arm
“Scientists at Caltech have developed an implantable brain chip that can translate the intentions of one tetraplegic man into movements of a robotic arm.” – Popular Mechanics
Check out the video here!
11. BRETT - Deep Learning for Robotics
This team of UC Berkeley researchers has developed algorithms that enable their robot, nicknamed BRETT for Berkeley Robot for the Elimination of Tedious Tasks, to learn new tasks through trial and error.
“The key is that when a robot is faced with something new, we won't have to reprogram it”
Check out some of the other robotics companies below working on a general purpose robot powered by AI.
From Rethink Robotics, meet Baxter– the safe, flexible, affordable alternative to outsourced labour and fixed automation.
From ABB Robotics, introducing Yumi - a collaborative, dual arm, small parts assembly robot solution that includes flexible hands, parts feeding systems, camera-based part location and state-of-the-art robot control.
And here’s a video compilation of robots falling down at the recent DARPA robotics challenge finals – a few things we can giggle at while AI-powered robotics is in the ‘deceptive’ phase of exponential technologies!
12. Google Wants to Patent its Creepiest Idea Ever
And finally, as suggested in my last AI blog referring to Google Panda – here is a creepy version of that April fools “prank” from Google that may hit the toy shop shelves one day.
So there have been quite few developments since our last blog focusing on AI.
At the end of 2014, Shivon Zilis from Bloomberg Beta put together a snapshot of the Machine Intelligence Landscape, where there has been an explosion of funding activity.
Despite the hype, there are definitely some real opportunities out there and many consider AI as almost ready for business. In addition, there are things that CEOs need to know to outwit the rise of AI. As discussed in Tech Trends 2015, Amplified Intelligence should play a dominant role in a post-Big Data world immersed by the Internet of Everything. However, emerging AI technologies will likely sit at the peak of inflated expectations in the next release of the Gartner Hype Cycle.
And so the AI arms race continues…
IBM acquired deep learning startup, AlchemyAPI for an undisclosed sum, adding 40,000 developers to its Bluemix platform. AlchemyAPI delivers a wide variety of text analysis and image recognition capabilities.
It will be interesting to see IBM’s response to the growing open source community as those deep learning projects begin to rival Watson’s capabilities. IBM’s $US 4 billion investment into cloud, big data/analytics, enterprise mobile/social and computer security does not take into account funding for further acquisitions. Meanwhile, startups like MetaMind continue to democratise the technology. MetaMind has been assembling quite the powerhouse team since it officially launched this year.
“Watson Health is our moonshot” – Ginni Rometty, CEO of IBM.
IBM went on to announce its Watson Health vertical and acquired two healthcare analytics companies, leveraging its relationships with Memorial Sloan-Kettering, MD Anderson and Emory University as well as partnering with Johnson & Johnson, Medtronic and Apple. The new business unit will have “at least 2000 consultants, medical practitioners, clinicians, developers and researchers to design, develop and accelerate the adoption of the Watson Health capability”.
With the launch of Apple Watch, IBM will aim to merge its health data with Apple, applying Watson to give users insights and advice from personal health information gathered from fitness trackers, smartphones, implants or other devices.
IP Australia will also trial Watson with the aim of improving agency efficiencies and enhancing online service offerings for Australia’s innovators.
And why not a cook book powered by Watson? Although the feedback has been mixed when it came to the Australian chocolate burrito…
Watson was also used to deliver answers at the Australian Open, based on uploaded historical data provided by Tennis Australia. These NAO robots were powered by Watson and are each sold separately!
Even the Kickstarter community is getting in on cognitive computing, funding Cognitoys - the internet-connected smart toys that learn and grow with a child. Which begs us to wonder whether Google Panda, one of Google’s many April Fools pranks this year, was actually market testing a potential product?
Google’s DeepMind revealed that it’s building a Neural Turing Machine. The result is a computer that mimics the short-term memory of the human brain.
More details surfaced from DeepMind’s work on an AI system that plays video games on its own.
“When the computers passed a level or racked up a high score, they were automatically rewarded with the digital equivalent of a dog treat. Google's AI system surpassed the performance of expert humans in 29 games, and outperformed the best-known algorithmic methods for completing games in 43 instances”- Demis Hassabis, DeepMind’s Co-Founder and VP of Engineering.
Hassabis said that the next step is to develop and train systems to navigate 3D worlds like in Tomb Raider. “If this can drive the car in a racing game, then potentially, with a few real tweaks, it should be able to drive a real car. That's the ultimate aim."
“Whoever wins artificial intelligence will win the Internet in China and around the world. Baidu has the best shot to make it work.”- Andrew Ng, Chief Scientist at Baidu said to Bloomberg.
Andrew Ng was one of the many AI researchers who attended the Re-Work Deep Learning Summit in San Francisco and shared his thoughts on AI development in a fireside chat.
Baidu currently holds the record for image recognition, havingfewer errors than Google and is close to human accuracy.
The Co-Founder and CEO of Baidu, Robin Li, has proposed a state-level project on AI - the “China Brain project” – and hopes to get support from the Chinese military. We’re in real arms race territory now!
Facebook acquired Wit.ai to help its developers with speech recognition and voice interfaces.
AI is one of three technologies that Facebook will focus on in its ten year plan – the other two being virtual reality and drones.
Yann LeCunn explained how Facebook tags your photos and what the technology holds in the future.
5. Nvidia and Autonomous Vehicles
Nvidia is betting big on AI– deep learning in particular. As we’ve learned, GPUs are much better than CPUs for running neural networks. CEO and Co-Founder of Nvidia, Jen-Hsun Huang appears to smile every time he hears more cameras will be installed in cars and pixels will be processed for autonomous vehicles during his interview with Elon Musk. It's been almost a decade since Nvidia began consciously pivoting away from its reputation as company that only makes products for PC gaming enthusiasts. It appears to be paying off in light of the slowdown in PC sales and growth in deep learning applications.
Check out the videos at NVIDIA’s conference for CES 2015 and you’ll see the role NVIDIA plays in the future of automotive technology and self-driving cars. There is an extensive pipeline for autonomous vehicles from all the major car companies. In this sense, self-driving cars are inevitable.
6. Other corporate AI developments
· Amazon has debuted a new 'Machine Learning' service during the AWS Summit in San Francisco.
· Cortana is coming to iOS and Android devices.
· Skype’s real-time translation tool now works across 4 languages – English, Spanish, Mandarin and Italian.
· H2O.ai, a leader in open source machine learning and deep learning, unveiled a new AI developer program designed to empower engineers with the tools to implement AI and Machine Learning and make their apps smarter.
7. AI in the mainstream
2015 is set to be quite a year for AI in the mainstream, with quite a few recent and upcoming blockbusters relating to the theme…
AI and robots even got its own episode on SBS One’s Insight – where the issues arising from automation, abundance, as well as existential risks were discussed.
8. Intellectual debate
So what is a robot anyway? And how could robotics and AI reach a level that could pose a threat to humanity?
A group of scientists and entrepreneurs, including Stephen Hawking and Elon Musk, have signed an open letter promising to ensure AI research benefits humanity. Shortly after, Bill Gates and Steve Wozniak joined Hawking and Musk in naming AI as one of humanity’s biggest existential risks.
If you’re still not sure what’s all the fuss – check out Nick Bostrom’s Superintelligence, where Bostrom argues that if machine brains surpass human brains in general intelligence, then this new superintelligence could replace humans as the dominant lifeform on Earth. Then there’s the realists working in the AI field who think those concerns should be left to Hollywood.
But rather than arguing whether to fear it or not, we should look at what advancements in AI mean for society, which is why Stanford University has launched a 100-year study on AI.
“AI today is advancing the diagnosis of disease, finding cures, developing renewable clean energy, helping to clean up the environment, providing high-quality education to people all over the world, helping the disabled…and contributing in a myriad of other ways. We have the opportunity in the decades ahead to make major strides in addressing the grand challenges of humanity. AI will be the pivotal technology in achieving this progress. We have a moral imperative to realize this promise while controlling the peril. It won’t be the first time we’ve succeeded in doing this.”
- Ray Kurzweil, Director of Engineering at Google said to Time.
Here’s that law of accelerating returns again from Ray Kurzweil.
The most immediate concern for society from AI, however, is job losses.
9. Staying competitive through enhancement
Some experts are confident that technology will enhance people’s biological capabilities to stay competitive with AI and automation in the face of rising unemployment over the next two decades.Virtual reality and augmented reality already allow us to work remotely or immerse ourselves with additional relevant information. In a recent Ted Talk, David Eagleman ask show can we create new senses for humans.
From smart contact lenses to synthetic eyes, we should expect to see a growing market for human enhancement through technology.
And then there are the biohackers whogave themselves night vision through eye drops…
But what if people decided to replace their limbs with prosthetics that function better than a natural one? Or connect the neurons in their brains to the internet? What, in the end, makes us human? In Ray Kurzweil’s book, The Singularity Is Near: When Humans Transcend Biology, he famously predicted that in 2045, humankind will have its Terminator moment: The rise of computers will outpace our ability to control them. To keep up, we will radically transform our biology via nanobots and other machines that will enhance our anatomy and our DNA, changing everything about how we live and die.
10. Life-Extension Research
“If you ask me today, is it possible to live to be 500? The answer is yes.” - Bill Marris, President and Managing Partner of Google Ventures.
In enhancing the human condition, Google Ventures appears to be on a search for immortality, backing research centres like Calico to study how to reverse aging. And Google’s not alone, with some of the biggest tech titans working on projects that can defy death. In a SXSW keynote, Martine Rothblatt discusses AI, immortality and the future of selves.
But would you want to live forever? Provided you stayed healthy, had your organs 3D printed for replacement and conquered any neurodegenerative disorders? With the advancement in medical research and drugs that are tailored to each person’s DNA, the possibility of living forever is starting to become a reality.
The media coverage on advancements in AI and interest from the tech titans on life-extension research has reinvigorated a movement coined in the 1950s as Transhumanism (or H+). Transhumanism is an international cultural and intellectual movement with an eventual goal of fundamentally transforming the human condition by developing and making widely available technologies to greatly enhance human intellectual, physical, and psychological capacities.
There’s now a Transhumanist Party formed in the US, with its leader,Zoltan Istvan (author of The Transhumanist Wager) running for President at the next US election. There is also a Transhumanist Party in the UK, led byAmon Twyman, which has anindependent running for the next UK election.
The transhumanist movement is picking up quite a bit of momentum and starting to have a strong physical presence around theworld. The policies of the Transhumanist Parties in each country are based on transhumanist values, taking into account where technology and transhumanism might take us. With automation comes joblessness and wealth inequality, so a Universal Basic Income is a common core policy of the Transhumanist Parties until society enters an age of abundance. Nevertheless, the culture of Transhumanism is about self-improvement.
12. Closing thoughts
In outlining all this, my hope is that topics such as automation, artificial super intelligence, transhumanism,mind uploading and the creation of acosmic consciousness might one day become a topic of mainstream discussion. Once exposed to these concepts, it’s hard to go back to watching ball sports and soap operas! Follow me on twitter @peterxing to stay posted on the latest AI developments.
1. Humans need not apply
As we’ve seen in No.8 of Sightings from the Edge August 13 (Pew Research Report, August 2014 - AI, Robotics and The Future of Jobs), experts envision automation and intelligent digital agents permeating vast areas of our work and personal lives by 2025, but they are divided on whether these advances will displace more jobs than they create.
There will inevitably be new job categories created by 2025 which are unfathomable today, but here’s a trending video from CPG Grey on why “humans need not apply” and how the AI revolution will have a completely different impact on humanity’s role in the workforce compared to the agricultural, industrial and digital revolutions which preceded it.
Here’s what the tech giants are doing to make this vision a reality, and the AI arms race that has ensued:
Since winning Jeopardy! over 3 years ago, IBM has finally made significant investment recently into opening up its Watson technology for developers and forming major alliances (most notably with Apple) to bring Watson to the masses. Watson works unlike any software processing code. Watson is a cognitive technology that processes information more like a human than a computer - by understanding natural language, it generates hypotheses based on evidence and it learns as it goes.
Here’s a 2 minute video that summarises what Watson is all about: What will you do with Watson?
So far, theapplications of Watson have found its way in medical diagnoses, business meetings rooms, BBQ sauce formulation, finance, civil transitions for military service members, and even in the legal profession. The possibilities are apparently endless – and unsurprisingly so, when the goal is to create a chip that replicates the human brain.
Stanford University professor, Andrew Ng started Google Brain (Google’s Deep Learning project) 3 years ago. By June 2012, Google Brain had learnt how to recognise cats in YouTube videos.
Earlier this year, Google acquired DeepMind, a company that has demonstrated its deep learning system can master games by simply playing them over and over again (here’s a demo of it mastering some retro games such as “Breakout”). As summarised by Digital Trends:
“It’s no secret that Google has an interest in artificial intelligence; after all, technologies derived from AI research help fuel Google’s core search and advertising businesses. AI also plays a key role in Google’smobile services, itsautonomous cars, and its growing stable of robotics technologies. And with the addition of futurist Ray Kurzweil to its ranks in 2012, Google also has the grandfather of ‘strong AI’ on board, a man who forecasts that intelligent machines may exist by mid-century”.
As mentioned in No. 4 of Sightings from the Edge July 4, Ray Kurzweil’s TED talk “Get Ready for Hybrid Thinking” emphasises the importance of acknowledging the law of accelerating returns. See Ray Kurzweil's website for cutting edge AI developments and his book, How to create a Mind: The Secret of Human Thought Revealed.
Google also recently acquired Jetpac, a company known for providing software that analyses websites for photos that are then compiled into city guides using AI. More ammunition for a smarter Google Now and other services.
Chinese tech company Baidu, which came up with its first ever visual search engine last year, has yet to make its popular search engine and other web services available in English. However, Baidu have been able to build a neural network that roughly matched the Google Brain system for a 50th of the cost ($20,000) using off-the-shelf graphics chips from Nvidia. Since hiring Andrew Ng from Google, Baidu is looking to apply neural networks which can work behind the scenes for a wide variety of applications, including those that handle text, spoken words, images, and videos.
Late last year, Facebook hired New York University professor Yann LeCun to run its new AI lab. LeCun developed a system in the 90’s that could recognise written digits. Automatically reading bank cheques, it marked the first time convolutional neural nets (convnets) were applied to practical problems. Now, LeCun is actively looking to hire more AI talent at the company, and it’s not difficult to see how Facebook’s accumulation of strategic assets will complement the data which will feed into Facebook’s convnets.
Mark Zuckerberg also has a significant stake in Vicarious, which cracked the CAPTCHA Turing test. Other notable investors in Vicarious include Elon Musk, Jerry Yang and Jeff Bezos.
As noted above, Apple’s alliance with IBM will also give Apple unprecedented access to enterprise.
However, its ex-Siri team has also been busy under the radar.
Meet Viv, Siri’s replacement. Viv is the next step in the evolution towards an OS envisioned in the movie “Her”.
“Whereas Siri can only perform tasks that Apple engineers explicitly implement, this new program, they say will be able to teach itself, giving it almost limitless capabilities. In time, they assert, their creation will be able to use your personal preferences and a near-infinite web of connections to answer almost any query and perform almost any function”.
Microsoft has developed a deep learning system modeled after the human brain that has greater image classification accuracy and is 50 times faster than other systems in the industry. It can successfully identify things like what breed of dog is in the picture you have taken on your phone.
Here’s a short video: “Introducing Project Adam: a new deep learning system from Microsoft”.
Combined with its Xbox Kinect depth camera technology and Cortana, it’s clear that Microsoft will be a key player in the AI arm’s race.
8. Other AI developments
· Last month, Twitter acquired deep learning startup Madbits to join the AI arms race. Madbits has a visual intelligence technology that automatically understands, organises and extracts relevant information from raw media.
· Even Spotify’s interns are working to apply deep learning to the business,building on convnets to create playlists based on certain features observed in the audio.
· And of course, start-ups will keep coming up with innovative ways to apply and invent new AI systems before they too, get bought up
9. Collective AI
Watch these miniature robots developed by researchers at Harvard University organise and communicate amongst themselves.
See more “Kilobots” in action.
It’s kind of scary to think what if these Kilobots shrunk to the size of molecules to become collectively intelligent, self-replicating nanobots?
10. Caution for the human race?
· Stephen Hawking warns AI could be a ‘real danger’
· Elon Musk is worried about AI, and thinks they are potentially “more dangerous than nukes”
· With the release of so many Hollywood dystopic blockbusters over the decades, it’s not hard to imagine how things may turn out for the human race:
Surely, there must be an off switch? What about Asimov’s 3 laws of robotics?
11. Merge with the machines?
· Many predict that within the lifetimes of people living today, we may be able to merge our consciousness with that of the machine. FutureTimeline.net is a speculative timeline of future history. Part fact and part fiction, the timeline is based on detailed research – including analysis of current trends, long-term environmental changes, advances in technology such as Moore's Law, future medical breakthroughs, the evolving geopolitical landscape and much more. Where possible, references have been provided to support the predictions.
· For a “mind-opening” read, check out theoretical physicist and futurist, Dr. Michio Kaku’s The Future of the Mind: The Scientific Quest to Understand, Enhance, and Empower the Mind. Dr Kaku predicts that in the not too distant future, we could upload our brain to a computer, neuron for neuron, and send our thoughts and emotions around the world on a “brain-net”.
12. Closing thoughts
We’ve come a long way since the hype leading up to the AI winter of the ‘80s and ‘90s. Now that computing power is catching up to the power of the human brain, the significant investment from public and private sectors around the world should only add momentum towards the singularity.
It’s hard not to get excited (and concerned) about what’s to come. However, as we’ve seen in the past, hype can often disappoint. As long as we don’t get that bad ending from AI: Artificial Intelligence (2001), then I’m all for it.