What you'll find below are some of the best articles on AI and robotics that were shared in the CTOdaily newsletter in 2018.
For more content like this delivered right to your inbox, head to buckhamduffy.com/subscribe.
This is a great piece on some of the conceptual and performance issues of AI. The author explores the current state of AI and how majority of these instances of artificial intelligence should instead be considered examples of 'computational statistics' (see: AI, Machine Learning and Data) rather than real intelligence. We still really have no idea how an information processing entity can make the jump from processing to thinking, so the author argues, there is good reason to expect human labour to be necessary for a very long time. That being said, there are still a huge number of jobs that these 'computational statistic machines' can automate.
Like anything intelligent, AI needs to be trained. This piece is a concise and informative explanation of why data is foundational to intelligent machine design and how it ties into machine learning.
Fast Robotics Limited from Western Australia has developed Hadrian X -- a mobile robot brick layer capable of laying down 1000 bricks per hour (a human worker can lay between 300-400 per day) in uncontrolled environments.
This piece is a bit older than the ones I normally share, but given the ubiquity of storytelling and its importance, I thought it would be good to highlight it here. The famous writer Kurt Vonnegut said that “there is no reason why the simple shapes of stories can’t be fed into computers. They are beautiful shapes.” It turns out he might have been right. His idea resonated with a group of mathematicians and computer scientists and they decided to build an experiment around it. With "sophisticated computing power, natural language processing, and reams of digitized text" they mapped the narrative patterns in a huge corpus of literature. What they found was that stories generally have the following emotional arcs:
1. Rags to Riches (rise)
2. Riches to Rags (fall)
3. Man in a Hole (fall then rise)
4. Icarus (rise then fall)
5. Cinderella (rise then fall then rise)
6. Oedipus (fall then rise then fall)
This is definitely something to keep in mind for any of you storytellers out there!
Artificial Intelligence is set to be the keystone technology of the coming decades as it enables us to outsource parts of our thinking and decision making, helping us make sense of and act upon the huge amounts of data we're creating. Puneet Shivam, president of Avendus Capital US, an investment bank, says: “The leaders in the AI cloud will become the most powerful companies in history.” This is partly due to the nature of machine learning—the more data the system gets, the better the decisions it will make. This means customers are more likely to get locked in to an initial vendor. The first company to deliver outstanding AI value could be the first trillion dollar company. Microsoft, Google, Amazon, Alibaba and Baidu are just some of the juggernauts all trying to become dominant in the AI space
In this piece, Andrew Ng, cofounder of Coursera, AI Fund, and Landing.AI, discusses the difference between an AI-enabled business versus a true AI company, and how businesses can organise, hire, and make use of AI to add value. He's a thought-leader in the space, and came up with a great analogy for building AI. “AI is akin to building a rocket ship. You need a huge engine and a lot of fuel. The rocket engine is the learning algorithms but the fuel is the huge amounts of data we can feed to these algorithms.”
Poor data quality is enemy number one to the widespread, profitable use of machine learning. The 'garbage-in, garbage-out' is all too true when it comes to machine learning. The bigger the difference between the data you possess and how well it maps onto reality, the worse your machine learning outcomes will be. According to the author of the piece, data concerns must be "met with an aggressive, well-executed quality program" which "requires the leaders of the overall effort to take all of the following five steps:"
- Clarify your objectives and assess whether you have the right data to support these objectives.
- Build plenty of time to execute data quality fundamentals into your overall project plan.
- Maintain an audit trail as you prepare the training data.
- Charge a specific individual (or team) with responsibility for data quality as you turn your model loose.
- Obtain independent, rigorous quality assurance.
"Legendary futurist and a director of engineering at Google, Ray Kurzweil, recently introduced 'Talk to Books' - a new way to find answers on the internet that should bring pleasure to researchers, bookworms and anyone seeking to expand their thinking on a range of topics. Type a question into 'Talk to Books,' and AI-powered tool will scan every sentence in 100,000 volumes in Google Books and generate a list of likely responses with the pertinent passage bolded."
JX Press Corp is a Japanese news technology venture that uses artificial intelligence to analyse social media to get news scoops before the major players. Their machine intelligence "scours social media postings, analyzing text, photos and even exclamation marks, to find breaking news in Japan, in areas such as fires, traffic accidents and other disasters. It also monitors overseas media and Twitter accounts that it considers trustworthy, seeking to be the first to report major international developments. Once it’s found news, its algorithms write the stories."
IBM recently hosted two debate club-style discussions between two humans and an AI called “Project Debater.” The aim was for the AI to engage in a series of reasoned arguments according to some standard rules of debate, like not knowing the topic ahead of time, a four-minute introductory speech, a four-minute rebuttal to the other’s arguments, and a two-minute closing statement. The AI held its own. According to the author, it "created a freshman-level term paper kind of argument in just a couple of minutes when presented with a debating topic it had no specific preparation for." "The system has several hundred million articles that it assumes are accurate in its data banks, which cover around 100 areas of knowledge. When it gets a debate topic, it takes a couple of minutes to go through them, decides what would make the best arguments in favour of the topic, and then creates a little speech describing those points." This is a thrilling development. It shows us that soon we may have digital partners to help us think and reason that can leverage absurd amounts of data. This is still a research-level project, though the technologies underneath it right now are already beginning to be used in IBM projects according to Jeff Welser, the VP and lab director for IBM research at Almaden.
OpenAI have defeated a team of former professionals in a number of games of Dota 2, one of the most watched and complex e-sports in the world. In may, the OpenAI bots were losing to amateur players."By June, the AI had matured enough to defeat casual players, and today it’s shown itself capable of overwhelming people who’ve been playing Dota 2 literally since its inception. The next goal for this rapidly evolving AI is to take on the very best Dota 2players at Valve’s The International 8 later this month." OpenAI have come such a long way in only a year. In August of last year they released a bot that could convincingly beat professional players in 1v1 game, orders of magnitude less complex than a 5v5 game which they recently dominated at.
AI is set to be the lynchpin technology of the information age, empowering us to analyse, interpret and act upon the ever-increasing amounts of data we're creating on a daily basis. This piece explores how AI will impact the world economy, exploring issues like inequality (within countries as well as within industries), how these changes will impact employees, and the need for policymakers to understand and adapt Here are some of the takeaways from the piece: AI could lead to a performance gap between companies within industries, as the benefits derived from these cognitive technologies could increase performance, creating a positive feedback loop. AI could help shift demand for jobs away from repetitive tasks toward those that are socially and cognitively driven and require more digital skills. Leading AI countries could capture an additional 20 to 25 percent in net economic benefits, compared with today, while developing countries might capture only about 5 to 15 percent.