Daily Tech Digest - December 19, 2020

The road to successful change is lined with trade-offs

Leaders should borrow an important concept from the project management world: Go slow to go fast. There is often a rush to dive in at the beginning of a project, to start getting things done quickly and to feel a sense of accomplishment. This desire backfires when stakeholders are overlooked, plans are not validated, and critical conversations are ignored. Instead, project managers are advised to go slow — to do the work needed up front to develop momentum and gain speed later in the project. The same idea helps reframe notions about how to lead organizational change successfully. Instead of doing the conceptual work quickly and alone, leaders must slow down the initial planning stages, resist the temptation and endorphin rush of being a “heroic” leader solving the problem, and engage people in frank conversations about the trade-offs involved in change. This does not have to take long — even just a few days or weeks. The key is to build the capacity to think together and to get underlying assumptions out in the open. Leaders must do more than just get the conversation started. They also need to keep it going, often in the face of significant challenges. 


AI, ML can bolster cybersecurity, and vice versa, professor says

Machine learning and artificial intelligence for cybersecurity. It's like two ways we are tackling the problems. For example, machine learning has good results in machine vision, computer vision, or you can look at the games. For example, chess was built that was using machine learning and artificial intelligence. As a result the chess game, played by computer, beat the smartest human a couple of years ago. And it has been a very promising application of AI and machine learning. So at the same time, you can look at how machine learning algorithms are compromised. If you recall, there was a Tesla car speeding. And I think it slipped by at 55 miles on the road [with a speed limit] of 35 miles per hour, just because of a smart piece of tape [on a sign].  In that case we are trying to use the advantage or the benefits that AI and machine learning offer for securing systems that we are envisioning in the years to come. But again, at the same time, can we use the other way around too, right. That's why my research is focusing on AI and machine learning for cybersecurity. At the same time, cybersecurity for AI because we want to secure the system that is working for a greater good. Another example that I could explain is if you are using, let's say a machine learning algorithm to filter out the applicants from the application pool to hire somebody.


Low Code: CIOs Talk Challenges and Potential

Former CIO Dave Kieffer says that “cloud ERP can't be customized, but they can be extended. CRM's can provide full platforms. Extension, not customization, should be the goal no matter the platform.” However, Sacolick suggests when “CIO can seamlessly, integrate data and workflow with ERP and CRM, then customization will be less needed. It can become an architecture decision that provides a lot more business flexibility. Low Code and No Code apps, integrations are one approach.” Deb Gildersleeve agrees and says, “a lot of these legacy systems can’t be customized, and for those that can be, most organizations don’t want to spend the time, money and resources needed to customize them. That’s where low code can come in to complement these systems and work within your existing tech stack.” Francis, meanwhile, suggests, “there will likely always be need for an amount of high coding. Sacolick says, “I call highcoding, DIYCoding. But the real challenge is getting app developers on board with using low code /no code where it makes sense. Many really love coding, and some lose sight that their role is to provide solutions and innovations.”


Where are we really with AI and ML? 5 takeaways from the AI Innovators Forum

The reality is AI and ML can’t be applied to every scenario. For example, one company created a sophisticated ML approach to examining marketing analytics and performance so AI could predict the most effective marketing channels, but it required an incredible number of data integrations, and it wasn’t predictably better at suggesting marketing channels than existing experts using their expertise.  Likewise it is not yet possible to train enough scenarios of kids running in the street or a car swerving into your lane for self driving cars to learn and act accordingly. What is the safe default? Should the car stop when it doesn’t know what to do? Should it revert to a manual mode? These are just some of the ongoing challenges in training the ‘last 10%” of AI despite the vast majority of driving decisions being more efficient and accurate than a human driving on the road today. Beyond the more obvious self-driving car scenarios, many AI and ML use cases bring ethical considerations that should be taken seriously. Like any other technology at scale, there must be frameworks and guardrails to help understand potential impact, mitigation paths, and when to forgo the use of these technologies altogether. 


Ensuring Data Residency and Portability

Even if the server is based in California and stores only Indian data, it does not come within the sovereign jurisdiction of India or Indian courts. While it may be technically possible to isolate a set of data within the server that is deemed perverse or critical evidence and electronically “seal” it, the California company might not be interested in blocking large space within its server or servers for which it has invested millions, for a court case in faraway India. This is just one example of the limits of national laws versus the limitless, borderless movements of data, made possible by technology. Hence, data localisation is an issue that has been a hot topic for governments around the world. Alongside comes the issue of data portability. What does the right to data portability mean? This is a right that allows anybody who has put a set of data into one service or site, to obtain it from that service or site and reuse it for their own purposes across different services. The sense of portability is in the moving, copying and transferring personal data from one to another service without compromising on security. This right will also incorporate the right to have the quality of data undiminished or unchanged. Within this right is also incorporated the caveat that all such data will have only been gathered from the user with his or her consent.


Countries that retaliate too much against cyberattacks make things worse for themselves

“If one country becomes more aggressive, then the equilibrium response is that all countries are going to end up becoming more aggressive,” says Alexander Wolitzky, an MIT economist who specializes in game theory. “If after every cyberattack my first instinct is to retaliate against Russia and China, this gives North Korea and Iran impunity to engage in cyberattacks.” But Wolitzky and his colleagues do think there is a viable new approach, involving a more judicious and well-informed use of selective retaliation. “Imperfect attribution makes deterrence multilateral,” Wolitzky says. “You have to think about everybody’s incentives together. Focusing your attention on the most likely culprits could be a big mistake.” The study is a joint project, in which Sandeep Baliga, the John L. and Helen Kellogg Professor of Managerial Economics and Decision Sciences at Northwestern University’s Kellogg School of Management added to the research team by contacting Wolitzky, whose own work applies game theory to a wide variety of situations, including war, international affairs, network behavior, labor relations, and even technology adoption.


Driving autonomous vehicles forward with intelligent infrastructure

As cars are becoming more autonomous, cities are becoming more intelligent by using more sensors and instruments. To drive this intelligence forward, smart city IT infrastructure must be able to capture, store, protect and analyze data from autonomous vehicles. Similarly, autonomous vehicles could greatly improve their performance by integrating data from smart cities. In smart city planning, stakeholders must consider how they will enable the sharing of data in both directions, to and from autonomous vehicles, and how that data can be analysed and acted upon in real-time, so traffic keeps moving and drivers, passengers and pedestrians are kept safe. This means that a city needs physical infrastructure to handle the growing numbers of autonomous vehicles that will be on the streets and an IT infrastructure that can easily manage data storage, performance, security, resilience, mobilisation and protection from a central management console. For example, there’s a case to be made that cities should already be building networks of smart sensors along the roadside. These would have the capability to measure traffic conditions and potentially even monitor obstacles such as fallen trees, traffic collisions or black ice.


How teaching 'future resilient' skills can help workers adapt to automation

Automation itself isn’t a problem, but without a reskilling strategy it will be. Here’s how quality non-degree credentials can help. This fear of automation is not new. As the late Harvard professor Calestous Juma laid out in his seminal book Innovation and Its Enemies: Why People Resist New Technologies, technological progress has always come with some level of public concern. The bellhops feared automatic elevators and so did bowling pin resetters. Video did indeed “kill the radio star” and it wasn’t long before internet media streaming services made video retailers obsolete in the mid-2000s. This “creative destruction” means that automation-enabling technologies will destroy jobs, but they will also increase productivity, lower prices and create new (hopefully better) jobs too. Some have even advocated that in order to help low-income workers, we should speed up the automation of low-income jobs. Non-degree credentials can help workers adapt. Automation can change the world for the better, but only we if prepare for it. To be sure, non-degree credentials are no silver bullet to automation displacement. A number of policy recommendations can help our world transition to new, high-quality jobs.


ML-Powered Digital Twin For Predictive Maintenance — Notes From Tiger Analytics

In the last decade, the Industrial Internet of Things (IIoT) has revolutionised predictive maintenance. Sensors record operational data in real-time and transmit it to a cloud database. This dataset then feeds a digital twin, a computer-generated model that mirrors the physical operation of each machine. The concept of the digital twin has enabled manufacturing companies not only to plan maintenance but to get early warnings of the likelihood of a breakdown, pinpoint the cause, and run scenario analyses in which operational parameters can be varied at will to understand their impact on equipment performance. Several eminent ‘brand’ products exist to create these digital twins, but the software is often challenging to customise, cannot always accommodate the specific needs of every manufacturing environment, and significantly increases the total cost of ownership. ML-powered digital twins can address these issues when they are purpose-built to suit each company’s specific situation. They are affordable, scalable, self-sustaining, and, with the right user interface, are extremely useful in telling machine operators the exact condition of the equipment under their care.


Artificial intelligence stands at odds with the goals of cutting greenhouse emissions. Here’s why

What does this mean for the future of AI research? Things may not be as bleak as they look. The cost of training might come down as more efficient training methods are invented. Similarly, while data center energy use was predicted to explode in recent years, this has not happened due to improvements in data center efficiency, more efficient hardware and cooling. There is also a trade-off between the cost of training the models and the cost of using them, so spending more energy at training time to come up with a smaller model might actually make using them cheaper. Because a model will be used many times in its lifetime, that can add up to large energy savings. In my lab’s research, we have been looking at ways to make AI models smaller by sharing weights or using the same weights in multiple parts of the network. We call these shapeshifter networks because a small set of weights can be reconfigured into a larger network of any shape or structure. Other researchers have shown that weight-sharing has better performance in the same amount of training time. Looking forward, the AI community should invest more in developing energy-efficient training schemes.



Quote for the day:

"Any man who has ever led an army, an expedition, or a group of Boy Scouts has sadism in his bones." -- Tahir Shah

No comments:

Post a Comment