tag:ai.economic-evolution.net,2013:/posts Economic Evolution's AI Blog 2019-01-16T18:30:18Z Edited by JPM tag:ai.economic-evolution.net,2013:Post/1363993 2019-01-16T18:24:00Z 2019-01-16T18:30:18Z WSJ Video: Japanese "Robot Hotel" Eliminates its Robots

This video shows the vision of hotel without human employees. But it did not work. 

The WSJ report on January 14, 2019. 

Turns out, robots aren’t the best at hospitality. After opening in a blaze of publicity in 2015, Japan’s Henn na, or “Strange,” Hotel, recognized by the Guinness Book of World Records as the world’s first robot hotel, is now laying off its low-performing droids. 

Please stop talking and let me sleep.

So far, the hotel has culled over half of its 243 robots, many because they created work rather than reduced it.

[...]

Mr. Sawada said he hasn’t given up on the idea of a hotel without human staff, but Strange Hotel has taught him that there are currently many jobs suited only for humans. “When you actually use robots you realize there are places where they aren’t needed—or just annoy people,” he said.


Full Story at WSJ.com. 

]]>
JPM
tag:ai.economic-evolution.net,2013:Post/1360434 2019-01-05T12:00:45Z 2019-01-05T12:00:45Z The Seven Deadly Sins of AI Predictions

This is a great article by Rodney Brooks, who is a former director of the Computer Science and Artificial Intelligence Laboratory at MIT and a founder of Rethink Robotics and iRobot. 

Excerpt: Similarly, we have seen a sudden increase in performance of AI systems thanks to the success of deep learning. Many people seem to think that means we will continue to see AI performance increase by equal multiples on a regular basis. But the deep-learning success was 30 years in the making, and it was an isolated event.

That does not mean there will not be more isolated events, where work from the backwaters of AI research suddenly fuels a rapid-step increase in the performance of many AI applications. But there is no “law” that says how often they will happen.

6. Hollywood scenarios

The plot for many Hollywood science fiction movies is that the world is just as it is today, except for one new twist.

In Bicentennial Man, Richard Martin, played by Sam Neill, sits down to breakfast and is waited upon by a walking, talking humanoid robot, played by Robin Williams. Richard picks up a newspaper to read over breakfast. A newspaper! Printed on paper. Not a tablet computer, not a podcast coming from an Amazon Echo–like device, not a direct neural connection to the Internet.

It turns out that many AI researchers and AI pundits, especially those pessimists who indulge in predictions about AI getting out of control and killing people, are similarly imagination-challenged. They ignore the fact that if we are able to eventually build such smart devices, the world will have changed significantly by then. We will not suddenly be surprised by the existence of such super-intelligences. They will evolve technologically over time, and our world will come to be populated by many other intelligences, and we will have lots of experience already. Long before there are evil super-intelligences that want to get rid of us, there will be somewhat less intelligent, less belligerent machines. Before that, there will be really grumpy machines. Before that, quite annoying machines. And before them, arrogant, unpleasant machines. We will change our world along the way, adjusting both the environment for new technologies and the new technologies themselves. I am not saying there may not be challenges. I am saying that they will not be sudden and unexpected, as many people think.

7. Speed of deployment

New versions of software are deployed very frequently in some industries. New features for platforms like Facebook are deployed almost hourly. For many new features, as long as they have passed integration testing, there is very little economic downside if a problem shows up in the field and the version needs to be pulled back. This is a tempo that Silicon Valley and Web software developers have gotten used to. It works because the marginal cost of newly deploying code is very, very close to zero.

Deploying new hardware, on the other hand, has significant marginal costs. We know that from our own lives. Many of the cars we are buying today, which are not self-driving, and mostly are not software-enabled, will probably still be on the road in the year 2040. This puts an inherent limit on how soon all our cars will be self-driving. If we build a new home today, we can expect that it might be around for over 100 years. The building I live in was built in 1904, and it is not nearly the oldest in my neighborhood.

Capital costs keep physical hardware around for a long time, even when there are high-tech aspects to it, and even when it has an existential mission.

The U.S. Air Force still flies the B-52H variant of the B-52 bomber. This version was introduced in 1961, making it 56 years old. The last one was built in 1962, a mere 55 years ago. Currently these planes are expected to keep flying until at least 2040, and perhaps longer — there is talk of extending their life to 100 years.

I regularly see decades-old equipment in factories around the world. I even see PCs running Windows 3.0 — a software version released in 1990. The thinking is “If it ain’t broke, don’t fix it.” Those PCs and their software have been running the same application doing the same task reliably for over two decades.

The principal control mechanism in factories, including brand-new ones in the U.S., Europe, Japan, Korea, and China, is based on programmable logic controllers, or PLCs. These were introduced in 1968 to replace electromechanical relays. The “coil” is still the principal abstraction unit used today, and PLCs are programmed as though they were a network of 24-volt electromechanical relays. Still. Some of the direct wires have been replaced by Ethernet cables. But they are not part of an open network. Instead they are individual cables, run point to point, physically embodying the control flow — the order in which steps get executed — in these brand-new ancient automation controllers. When you want to change information flow, or control flow, in most factories around the world, it takes weeks of consultants figuring out what is there, designing new reconfigurations, and then teams of tradespeople to rewire and reconfigure hardware. One of the major manufacturers of this equipment recently told me that they aim for three software upgrades every 20 years.

In principle, it could be done differently. In practice, it is not. I just looked on a jobs list, and even today, this very day, Tesla Motors is trying to hire PLC technicians at its factory in Fremont, California. They will use electromagnetic relay emulation in the production of the most AI-enhanced automobile that exists.

A lot of AI researchers and pundits imagine that the world is already digital, and that simply introducing new AI systems will immediately trickle down to operational changes in the field, in the supply chain, on the factory floor, in the design of products.

Nothing could be further from the truth. Almost all innovations in robotics and AI take far, far, longer to be really widely deployed than people in the field and outside the field imagine.

Read Full Article from MIT Technology Review

]]>
JPM
tag:ai.economic-evolution.net,2013:Post/1342719 2018-11-11T08:46:22Z 2018-11-11T08:47:22Z Consulting companies getting active in AI

Alert: The link below is an advertisement by Accenture.  It was co-produced by the WSJ (Dow Jones) advertisement department. 

The only reason why we listing it here because it is interesting to see how consulting firms are trying to get into the game of advising companies on they should respond to the rise of machine learning. 

Accenture ad on WJS.com produced by "WSJ Studios"




]]>
JPM
tag:ai.economic-evolution.net,2013:Post/1335607 2018-10-24T17:36:08Z 2018-10-24T17:36:09Z Leslie Wilcox (LSE): AI will not cause net job losses
This video summarizes the conclusion reached in his book "Robotic Process and Cognitive Automation"
]]>
JPM
tag:ai.economic-evolution.net,2013:Post/1324690 2018-09-23T09:39:57Z 2018-10-24T17:39:58Z Interesting Essay on AI by the Ex-head of Google China

The Human Promise of the AI Revolution

Artificial intelligence will radically disrupt the world of work, but the right policy choices can make it a force for a more compassionate social contract.

Here are a couple of exerpts: 

"If handled with care and foresight, this AI crisis could present an opportunity for us to redirect our energy as a society to more human pursuits: to taking care of each other and our communities. To have any chance of forging that future, we must first understand the economic gauntlet that we are about to pass through.

Many techno-optimists and historians would argue that productivity gains from new technology almost always produce benefits throughout the economy, creating more jobs and prosperity than before. But not all inventions are created equal. Some changes replace one kind of labor (the calculator), and some disrupt a whole industry (the cotton gin). Then there are technological changes on a grander scale. These don’t merely affect one task or one industry but drive changes across hundreds of them. In the past three centuries, we’ve only really seen three such inventions: the steam engine, electrification and information technology.

[...]

AI’s main advantage over humans lies in its ability to detect incredibly subtle patterns within large quantities of data and to learn from them. While a human mortgage officer will look at only a few relatively crude measures when deciding whether to grant you a loan (your credit score, income and age), an AI algorithm will learn from thousands of lesser variables (what web browser you use, how often you buy groceries, etc.). Taken alone, the predictive power of each of these is minuscule, but added together, they yield a far more accurate prediction than the most discerning people are capable of.

For cognitive tasks, this ability to learn means that computers are no longer limited to simply carrying out a rote set of instructions written by humans. Instead, they can continuously learn from new data and perform better than their human programmers. For physical tasks, robots are no longer limited to repeating one set of actions (automation) but instead can chart new paths based on the visual and sensor data they take in (autonomy). 

Together, this allows AI to take over countless tasks across society: driving a car, diagnosing a disease or providing customer support. AI’s superhuman performance of these tasks will lead to massive increases in productivity. According to a June 2017 study by the consulting firm PwC, AI’s advance will generate $15.7 trillion in additional wealth for the world by 2030. This is great news for those with access to large amounts of capital and data. It’s very bad news for anyone who earns their living doing soon-to-be-replaced jobs."




]]>
JPM
tag:ai.economic-evolution.net,2013:Post/1324531 2018-09-22T21:53:34Z 2018-09-22T21:53:34Z AI is finding its way into hiring of big companies

This is a very interesting report by the WSJ. 

]]>
JPM
tag:ai.economic-evolution.net,2013:Post/1319996 2018-09-09T17:20:59Z 2018-09-09T17:20:59Z 46 Corporations Working On Autonomous Vehicles

Source: CB Insights

]]>
JPM
tag:ai.economic-evolution.net,2013:Post/1311311 2018-08-12T12:30:02Z 2018-08-12T12:30:02Z So Far AI has not had a big impact on Medical Diagnosis even after IBM spent billions

Can artificial intelligence lead to better cancer treatments? IBM spent six years and billions of dollars trying to find out.  

WSJ reports: "Oncology won’t be “a great space for making [AI] products” until there’s better data about patients, spanning genetic, environmental, lifestyle and health information, said Bob Kocher, a medical doctor and partner at venture-capital firm Venrock in Palo Alto, Calif. In the near term, most of the benefits from AI in the health-care field will come in administrative tasks such as billing, he added.




]]>
JPM
tag:ai.economic-evolution.net,2013:Post/1287882 2018-05-26T22:49:54Z 2018-05-26T22:49:54Z "To Build Truly Intelligent Machines, Teach Them Cause and Effect"

Judea Pearl on the future of AI: 

"The key, he argues, is to replace reasoning by association with causal reasoning. Instead of the mere ability to correlate fever and malaria, machines need the capacity to reason that malaria causes fever. Once this kind of causal framework is in place, it becomes possible for machines to ask counterfactual questions — to inquire how the causal relationships would change given some kind of intervention — which Pearl views as the cornerstone of scientific thought. Pearl also proposes a formal language in which to make this kind of thinking possible — a 21st-century version of the Bayesian framework that allowed machines to think probabilistically..."


Full Story


]]>
JPM
tag:ai.economic-evolution.net,2013:Post/1282365 2018-05-11T09:14:30Z 2018-05-11T09:14:59Z What meetings of the future might be like thanks to AI ]]> JPM tag:ai.economic-evolution.net,2013:Post/1274390 2018-04-19T00:03:00Z 2018-10-24T17:42:02Z AI Bookshelf

Here we will post AI books that we find worthy of attention. 


]]>
JPM
tag:ai.economic-evolution.net,2013:Post/1270215 2018-04-08T11:11:42Z 2018-04-08T11:11:42Z CEO of Microsoft on "Turning artificial intelligence into value"

This is the excerpt a McKinsey Quarterly (April 2018) interview with Microsoft CEO Satya Nadella. 


Simon London (Mckinsey Quarterly): Can we pivot and talk a little about AI? What advice do you give to executives that you talk to about how to leverage AI in their businesses?

Satya Nadella (CEO Microsoft): I believe AI is one of the more defining technologies of our time. One of the things I am most excited about is AI technology helping with inclusivity. For example, in the latest release of Windows, we have something called Eye Gaze, which allows anybody who is suffering from ALS [amyotrophic lateral sclerosis] to be able to type just with their gaze. We have learning tools inside of Word and OneNote that allow anyone with dyslexia to improve their reading. It’s powerful stuff, and it’s a very practical way for executives to deploy some of these tools so that more people in their workforces can fully participate, which is important.

But there is no question that automation and the efficiencies of automation are tremendously important. For example, if you go to support.microsoft.com, it’s a bot. It uses some of the latest techniques of reinforcement learning to answer questions that customers may have. And of course, if it runs out of gas, it turns over to the customer-service representative, who is also using the bot to help answer the question.

So, we have the full gamut of technology that is getting deployed. We now really have human-level speech recognition. In January, there was a contest at Stanford University for machine reading and comprehension. Microsoft was number one. This means a machine can read a piece of text and start answering questions, like a reading-comprehension test, without necessarily being fed the answers that are indexed in the text.

The advances are enormous, and they will lead to productivity gains broadly. Therefore, every CEO—every executive—should be thinking about how to get more analytical power or predictive power inside his or her business process or organization. That’s ultimately what’s needed to translate AI capability into productivity.

Simon London: What types of IT configuration and IT capabilities do companies need to do this?

Satya Nadella: I would say there are two big considerations. One of the fundamental things is that there’s no way to create AI if you don’t have data. If the data inside your organization is siloed, it’s going to be a challenge to create AI. This goes back to your point around company configuration.

Take customer connection as an example. In order to be much better at omnichannel customer connection—and it doesn’t matter whether you’re a retailer, a CPG [consumer-packaged-goods] company, or a bank—everything from the log data from your website, to your mobile analytics, to your CRM [customer-relationship-management] system, to all the other data streams, it all has to come together in order to create the next best touchpoint action with the customer. This is both an AI problem and a data problem. One of the things that we like to stress is: how can we help our customers first get their data estate in many cases into the cloud? Then they can reason on top of it and create these transformative outcomes, whether it is connecting with customers, or operational efficiencies, or even changing the nature of their products. This is a super important thing.

I would also add that trust is going to be of paramount importance. Not just the security side of trust but also the trust of the business model. You need to pick partners who are going to help you with your capability building, whose interests are aligned with your interests in the long term.

Simon London: If you’re a senior executive at a big industrial company, for example, there are a lot of different potential use cases for AI. Do you have any generalizable advice about how to look across those use cases and what to go for first?

Satya Nadella: When I pattern match and look at some of the best and easy-to-get-started use cases, it would be anything related to customer experience. This is a good use case. Let’s say there’s omnichannel customer data. The ability to do the next best action, whether it be a sales force, or inside sales, or your website personalization, can come in a variety of different ways.

Connecting with your customers more deeply—using your data and your ability to reason over data—using the latest AI techniques is one use case. The second use case is supply-chain or operational efficiencies. The IoT [the Internet of Things] is a fascinating thing. If you think about it, most of these projects are where you have a good or a service, you’re collecting operational data from it, you’re doing preventive maintenance, and then you’re going to connect it to field service, because once you can predict something, you want to connect it to somebody coming and fixing it before it’s broken. That’s a thing that can drive both top-line and bottom-line efficiencies. That’s a great use case, and we see a lot of it, especially in industrial companies. We also see a lot of deployment of technology to empower people inside the organization. I’m fascinated to see how HoloLens is being used for doing oil-field inspections or training. So, AI can be deployed not just against traditional knowledge work but also in what I will call frontline work.

Sometimes organizations have this “cobbler’s children” problem. They talk about all these great things they’ll do for customers, for [business] partners. Except you also need to do fantastic things for your employees so that they can do all these great things for customers and partners.


Full article entitled "Microsoft's next act" 

]]>
JPM
tag:ai.economic-evolution.net,2013:Post/1262267 2018-03-25T06:12:00Z 2018-03-27T08:49:07Z We need to Watch Artificial Intelligence Last year MIT technology review published a meta-analysis when artificial intelligence  (AI) experts expected certain categories of jobs to be taken away by AI. 
If these estimates are true, there is a fundamental question for society: Will the economy be able to create new jobs fast enough to compensate for the loss of jobs due to AI. Economic history has shown that western capitalist societies have always created new jobs in the face of rapid technological change. 

But what if this time is different? 

The goal of Economic-evolution AI blog is to watch the progress of AI and help interpret what future scenarios are more and less likely to unfold.  We want to help develop an informed understanding whether the AI optimists or pessimists are more likely to be correct. 

Among the optimists are many economist historians such as Joel Mokyr (Interview) or  Caspar Hirschi (Interview in German)  who take the position that there is nothing to worry about. Societies will not have to change radically.  
The pessimists fear that AI will destroy jobs and unprecedented magnitude and speed. If the rate of job destruction is going to be far larger the rate of new job creation, then we will have to radically redesign society because jobs will no longer provide the money to pay livelihood of most people as is the case today. Aija Leiponen is an example of the pessimistic voices. 

Our goal is to watch very closely the evidence and help everyone make up their minds how optimistic or pessimistic we should be. And if society needs to be redesigned, we better start thinking and experimenting with new forms of supporting the livelihood of all members of society without relying mainly on paid jobs. 


]]>
JPM