Mediaboss Marketing

After decades of extravagant promises and frustrating disappointments, artificial intelligence(AI) is finally starting to deliver real-life benefits to early-adopting companies. Retailers on the digital frontier rely on AI-powered robots to run their warehousesand even to automatically order stock when inventory runs low. Utilities use AI to forecast electricity demand. Automakers harness the technology in self-driving cars.

A confluence of developments is driving this new wave of AI development. Computer power is growing, algorithms and AI models are becoming more sophisticated, and, perhaps most important of all, the world is generating once-unimaginable volumes of the fuel that powers AIdata. Billions of gigabytes every day, collected by networked devices ranging from web browsers to turbine sensors.

The entrepreneurial activity unleashed by these developments drew three times as much investment in 2016between $26 billion and $39 billionas it did three years earlier. Most of the investment in AI consists of internal R&D spending by large, cash-rich digital-native companies like Amazon, Baidu, and Google.

For all of that investment, much of the AI adoption outside of the tech sector is at an early, experimental stage. Few firms have deployed it at scale. In a McKinsey Global Institute discussion paper, Artificial intelligence: The next digital frontier?, which includes a survey of more than 3,000 AI-aware companies around the world, we find early AI adopters tend to be closer to the digital frontier, are among the larger firms within sectors, deploy AI across the technology groups, use AI in the most core part of the value chain, adopt AI to increase revenue as well as reduce costs, and have the full support of the executive leadership. Companies that have not yet adopted AI technology at scale or in a core part of their business are unsure of a business case for AI or of the returns they can expect on an AI investment.

However, early evidence suggests that there is a business case to be made, and that AI can deliver real value to companies willing to use it across operations and within their core functions. In our survey, early AI adopters that combine strong digital capability with proactive strategies have higher profit margins and expect the performance gap with other firms to widen in the next three years.

This adoption pattern is widening a gap between digitized early adopters and others. Sectors at the top of MGIs Industry Digitization Index, such as high tech and telecoms or financial services, are also leading AI adopters and have the most ambitious AI investment plans. These leaders use multiple technologies across multiple functions or deploy AI at the core of their business. Automakers, for example, use AI to improve their operations as well as develop self-driving vehicles, while financial-services companies use it in customer-experience functions. As these firms expand AI adoption and acquire more data, laggards will find it harder to catch up.

Click on the + symbols for more

Governments also must get ahead of this change, by adopting regulations to encourage fairness without inhibiting innovation and proactively identifying the jobs that are most likely to be automated and ensuring that retraining programs are available to people whose livelihoods are at risk from AI-powered automation. These individuals need to acquire skills that work with, not compete against, machines.

The future of AI will be innovative, but may not be shared equally. Companies based in the United States absorbed 66 percent of all external investments into AI companies in 2016, according to our global review; Chinawas second, at 17 percent, and is growing fast. Both countries have grown AI ecosystemsclusters of entrepreneurs, financiers, and AI usersand have issued national strategic plans in the past 18 months with significant AI dimensions, in some cases backed up by billions of dollars of AI-funding initiatives. South Korea and the United Kingdom have issued similar strategic plans. Other countries that desire to become significant players in AI would be wise to emulate these leaders.

Significant gains are there for the taking. For many companies, this means accelerating the digital-transformation journey. AI is not going to allow companies to leapfrog getting the digital basics right. They will have to get the right digital assets and skills in place to be able to effectively deploy AI.

Download the discussion paper on which this article is based, Artificial intelligence: The next digital frontier? (PDF3MB).

Here is the original post:
How artificial intelligence can deliver real value to ...

Can common business sense be programmed into AI? Many are certainly trying to do just that. But there are decisions that often require a level of empathy -- let alone common sense -- that may be too difficult to embed into algorithms. In addition, while AI and machine learning are the hot tickets of the moment, technologists and decision makers need to think about whether it offers a practical solution to every problem or opportunity.

These points came up at a panel at the recentAI Summit, in which participants agreed that AI shouldn't be considered the default solution to every business situation that arises. (I co-chaired the conference and moderated the panel.) For starters, AI is still a relatively immature technology, saidDrew Scarano, a panelist at the session and vice president of global financial services atAntWorks. "We might be too reliant on this technology, forgetting about the humans in the loop and how they play an integral part in complementing artificial intelligence in order to get desired results."

AI is being used for many purposes across all industries, but the risk is in de-humanizing the interpersonal qualities that help build and sustain companies. "Today we can use AI for anything from approving a credit card to approving a mortgage to approving any kind of lending vehicle," said Scarano. "But without human intervention to be able to understand there's more to a human than a credit score, there's more to a person than getting approved or denied for a mortgage."

Scarano poo-poos the notion that AI systems comprise anything close to a "digital workforce," noting that "it's just a way to sell more stuff. I can sell 50 digital workers rather than one system. But digital workforce is just a bunch of code that does a specific task, and that task can be repeatable, or be customized." Another panelist,Rod Butters, chief technology officer forAible, agrees, noting that "at the end of the day, it's a machine. In the end, it's all 1s and 0s." The way to make AI more in tune with the business "is to get better tooling, craft, and experience with applying these machines in ways that first and foremost is transparent, and secondly understandable in some way, and ultimately something that is achieving an outcome that is business oriented or community oriented."

AI may be able to deliver fine-grained results based on logic beyond the capacity of human brains, but this may actually "run counter to what the business needs to be doing strategically," says Butters. "Because you can't have the visibility, you get unintended consequences, which can lead to complete disparities and equity in the application of processes to your customer base." Importantly, "there needs to be a feedback loop to ensure solutions you're implementing are resonating with your customers, and they're enjoying the experience as much as you're enjoying creating the experience," according to panelistRobert Magno, solutions architect withRun:AI.

Other experts across the industry also voice concern that AI is being pushed too hard in ways in may not be needed. "AI is not the solution to every business problem," says Pieter Buteneers, director of engineering in machine learning and AI at Sinch. "It sounds sexy, but there are going to be times when it's better to lean into how to best address customer needs rather than blindly investing in new technology."

While AI has the potential to make business processes more efficient and affordable, "at the end of the day, it is still a machine," Buteneers says. "AI lacks human emotion and common sense, so it can make certain mistakes that humans, instinctively, would not. AI can be easily fooled in certain ways that humans would spot from a mile away. For those who worry that AI will replace human jobs, we invariably need people working alongside AI bots to keep them in check and maintain a human touch in business."

AI initiatives "must be aligned with the company's operational needs and workflows to ensure a high level of adoption," agrees Sameer Khanna, senior vice president of engineering at Pager. "Identifying real world problems with user feedback is essential. Once the product is rolled out, there must be a continuous effort to engage users, monitor performance and improve solutions over time."

There are areas worth exploring with AI, however. For instance, "AI can reach and even surpass human performance in strictly defined tasks such as image recognition and language understanding," Buteneers says. "Harnessing the power of natural language processing enables AI systems to understand, write and speak languages like humans do. This offers tremendous benefits for businesses -- deploying an NLP-equipped chatbot or voicebot to complement the work of live service agents, for example, frees up those live agents to respond to complicated inquiries that require a more human approach."

Buteneers notes that "breakthroughs in NLP are making an enormous difference in how AI understands input from humans. I've helped design chatbots that can now understand 100+ languages at once, with AI assistants that can search for answers within any given body of text. AI can even make live customer service agents more effective by reading along during a conversation and offering them suggested responses based on previous conversations, customer context or from a larger knowledge base. Different algorithms in the NLP field can identify and analyze a message that may be fraudulent, which can allow organizations to weed out any spam messages before they get sent to consumers. The applications of NLP can provide countless benefits to any business: it can help save time and money, enhance the customer experience, and automate processes."

Still, human oversight is essential to ensuring these solutions serve customers. "Reviewing AI results should be the standard design process of algorithms -- it's ignorant to believe that once you've set up your model, your job is done," Buteneers says.

Khanna relates how his own company's ideas for AI projects "come primarily from collaboration between our data scientists and internal and client business stakeholders." This partnership "generates well-defined and feasible AI projects that are grounded in business realities," he adds. "Our data engineers, data scientists, and machine learning engineers then implement these projects using open-source technologies and proprietary products from cloud providers."

Read the original here:
The never-ending effort to bake common business sense into artificial intelligence - ZDNet

Meta founder Mark Zuckerberg was met with a mixed reaction after touting his company's "exciting breakthrough" towards creating an artificial intelligence system that thinks "the way we do."

In a Facebook post on Thursday, Zuckerberg hailed the development of Meta's data2vec, a new artificial intelligence algorithm that is capable of learning about several different types of information without supervision. Zuckerberg predicted that the development could eventually be used to more effectively help people perform common tasks like cooking.

"Exciting breakthrough: Meta AI research built a system that learns from speech, vision and text without needing labeled training data," Zuckerberg wrote in the post. "People experience the world through a combination of sight, sound and words, and systems like this could one day understand the world the way we do."

"This will all eventually get built into AR glasses with an AI assistant so, for example, it could help you cook dinner, noticing if you miss an ingredient, prompting you to turn down the heat, or more complex tasks," he added.

See posts, photos and more on Facebook.

Although previous artificial intelligence systems have also used self-supervised learning, they have not been able to learn more than one type of information effectivelyfor example, a system that is able to decipher text effectively may be unable to interpret information from images.

A blog post from Meta developers described data2vec as "the first high-performance self-supervised algorithm that works for multiple modalities." The developers noted that the data2vec performed better than multiple single-use algorithms and said that it "brings us closer to building machines that learn seamlessly about different aspects of the world around them."

"The core idea of this approach is to enable AI to learn more generally: AI should be able to learn to do many different tasks, including ones that are entirely unfamiliar," a Meta spokesperson said in a statement to Newsweek. "We want a machine to not only recognize animals shown in its training data but also adapt to recognize new creatures in an environment if we tell it what they look like."

"The hope is that algorithms like this one will lead to powerful multi-modal, self-learning AI models," the spokesperson continued. "Meaning AI that can make sense of the physical and virtual worlds around us using all the senses that humans do simultaneously."

While many responded to Zuckerberg's post by congratulating him and sharing in his excitement about the potential applications of data2vec, others responded to his post by expressing fears that the "creepy" development could lead to a "nightmarish dystopia."

"The potential benefits of this are far outweighed by the unimaginable nightmarish dystopia it will create," Facebook user Brendon Shapiro wrote in response to Zuckerberg's post. "If I forget the lemon zest, well, it'll just have to be ok."

"I've said this loads of times before.. I'm getting Skynet vibes.. we've all seen The Terminator," wrote actor Ritchi Edwards, referring to the film franchise's fictional artificial intelligence network that becomes sentient and begins to attack humanity.

"Ummmm... this kinda sounds creepy," Facebook user Rachel Miller wrote. "I am literally one of Facebooks most vocal fans... but this... an intuitive Alexa?? I don't know if I want it telling me to pick up the socks on the floor or telling me to add more cinnamon..."

Meta's new development does bring artificial intelligence closer to the goal of replicating human-like learning and thinking. However, the algorithm is still far removed from the creation of an autonomous system that could represent any kind of realistic threat to people if left unchecked.

Update 01/21/22, 6:50 p.m. ET:This article has been updated to include a statement from a Meta spokesperson.

The rest is here:
Frightening Reality of Meta-Built Artificial Intelligence That Can Think 'the Way We Do' - Newsweek

While mammograms are currently the gold standard in breast cancer screening, swirls of controversy exist regarding when and how often they should be administered. On the one hand, advocates argue for the ability to save lives: Women aged 60-69 who receive mammograms, for example, have a 33 percent lower risk of dying compared to those who dont get mammograms. Meanwhile, others argue about costly and potentially traumatic false positives: A meta-analysis of three randomized trials found a 19 percent over-diagnosis rate from mammography.

Even with some saved lives, and some overtreatment and overscreening, current guidelines are still a catch-all: Women aged 45 to 54 should get mammograms every year. While personalized screening has long been thought of as the answer, tools that can leverage the troves of data to do this lag behind.

This led scientists from MITs Computer Science and Artificial Intelligence Laboratory (CSAIL) and Jameel Clinic for Machine Learning and Health to ask: Can we use machine learning to provide personalized screening?

Out of this came Tempo, a technology for creating risk-based screening guidelines. Using an AI-based risk model that looks at who was screened and when they got diagnosed, Tempo will recommend a patient return for a mammogram at a specific time point in the future, like six months or three years. The same Tempo policy can be easily adapted to a wide range of possible screening preferences, which would let clinicians pick their desired early-detection-to-screening-cost trade-off, without training new policies.

The model was trained on a large screening mammography dataset from Massachusetts General Hospital (MGH), and was tested on held-out patients from MGH as well as external datasets from Emory, Karolinska Sweden, and Chang Gung Memorial hospitals. Using the teams previously developed risk-assessment algorithm Mirai, Tempo obtained better early detection than annual screening while requiring 25 percent fewer mammograms overall at Karolinska. At MGH, it recommended roughly a mammogram a year, and obtained a simulated early detection benefit of roughly four-and-a-half months better.

By tailoring the screening to the patient's individual risk, we can improve patient outcomes, reduce overtreatment, and eliminate health disparities, says Adam Yala, a PhD student in electrical engineering and computer science, MIT CSAIL affiliate, and lead researcher on a paper describing Tempo published Jan. 13 in Nature Medicine. Given the massive scale of breast cancer screening, with tens of millions of women getting mammograms every year, improvements to our guidelines are immensely important.

Early uses of AI in medicine stem back to the 1960s, where many refer to the Dendral experiments as kicking off the field. Researchers created a software system that was considered the first expert kind that automated the decision-making and problem-solving behavior of organic chemists. Sixty years later, deep medicine has greatly evolved drug diagnostics, predictive medicine, and patient care.

Current guidelines divide the population into a few large groups, like younger or older than 55, and recommend the same screening frequency to all the members of a cohort. The development of AI-based risk models that operate over raw patient data give us an opportunity to transform screening, giving more frequent screens to those who need it and sparing the rest, says Yala. A key aspect of these models is that their predictions can evolve over time as a patients raw data changes, suggesting that screening policies need to be attuned to changes in risk and be optimized over long periods of patient data.

Tempo uses reinforcement learning, a machine learning method widely known for success in games like Chess and Go, to develop a policy that predicts a followup recommendation for each patient.

The training data here only had information about a patients risk at the time points when their mammogram was taken (when they were 50, or 55, for example). The team needed the risk assessment at intermediate points, so they designed their algorithm to learn a patients risk at unobserved time points from their observed screenings, which evolved as new mammograms of the patient became available.

The team first trained a neural network to predict future risk assessments given previous ones. This model then estimates patient risk at unobserved time points, and it enables simulation of the risk-based screening policies. Next, they trained that policy, (also a neural network), to maximize the reward (for example, the combination of early detection and screening cost) to the retrospective training set. Eventually, youd get a recommendation for when to return for the next screen, ranging from six months to three years in the future, in multiples of six months the standard is only one or two years.

Lets say Patient A comes in for their first mammogram, and eventually gets diagnosed at Year Four. In Year Two, theres nothing, so they dont come back for another two years, but then at Year Four they get a diagnosis. Now there's been two years of gap between the last screen, where a tumor could have grown.

Using Tempo, at that first mammogram, Year Zero, the recommendation might have been to come back in two years. And then at Year Two, it might have seen that risk is high, and recommended that the patient come back in six months, and in the best case, it would be detectable. The model is dynamically changing the patients screening frequency, based on how the risk profile is changing.

Tempo uses a simple metric for early detection, which assumes that cancer can be caught up to 18 months in advance. While Tempo outperformed current guidelines across different settings of this assumption (six months, 12 months), none of these assumptions are perfect, as the early detection potential of a tumor depends on that tumor's characteristics. The team suggested that follow-up work using tumor growth models could address this issue.

Also, the screening-cost metric, which counts the total screening volume recommended by Tempo, doesn't provide a full analysis of the entire future cost because it does not explicitly quantify false positive risks or additional screening harms.

There are many future directions that can further improve personalized screening algorithms. The team says one avenue would be to build on the metrics used to estimate early detection and screening costs from retrospective data, which would result in more refined guidelines. Tempo could also be adapted to include different types of screening recommendations, such as leveraging MRI or mammograms, and future work could separately model the costs and benefits of each. With better screening policies, recalculating the earliest and latest age that screening is still cost-effective for a patient might be feasible.

Our framework is flexible and can be readily utilized for other diseases, other forms of risk models, and other definitions of early detection benefit or screening cost. We expect the utility of Tempo to continue to improve as risk models and outcome metrics are further refined. We're excited to work with hospital partners to prospectively study this technology and help us further improve personalized cancer screening, says Yala.

Yala wrote the paper on Tempo alongside MIT PhD student Peter G. Mikhael, Fredrik Strand of Karolinska University Hospital, Gigin Lin of Chang Gung Memorial Hospital, Yung-Liang Wan of Chang Gung University, Siddharth Satuluru of Emory University, Thomas Kim of Georgia Tech, Hari Trivedi of Emory University, Imon Banerjee of the Mayo Clinic, Judy Gichoya of the Emory University School of Medicine, Kevin Hughes of MGH, Constance Lehman of MGH, and senior author and MIT Professor Regina Barzilay.

The research is supported by grants from Susan G. Komen, Breast Cancer Research Foundation, Quanta Computing, an Anonymous Foundation, the MIT Jameel-Clinic, Chang Gung Medical Foundation Grant, and by Stockholm Lns Landsting HMT Grant.

Original post:
Seeing into the future: Personalized cancer screening with artificial intelligence - MIT News

Artificial Intelligence smart machines able to learn how to carry out tasks and become increasingly good at them is everywhere in work today, and will only be more ubiquitous tomorrow.

3 Ways That Artificial Intelligence (AI) Will Change Your Job Forever

In fact Googles CEO Sundar Pichai recently predicted that it will turn out to be the most profound human invention so far more so that electricity, the internet, or even fire!

Certainly, I believe it has the potential to deeply impact everything about the way we live our lives, from how we travel, to how we connect and communicate with friends, and most definitely the way we work and do business.

Whatever job you do now, if it isnt affected by AI already, its very likely that it will be at some point in the not-so-distant future. Heres my rundown of the five most significant changes AI will make to the world of work in our lifetimes.

AI probably wont make you redundant yet!

Its certainly true that machine learning the AI technology thats most relevant to business today will be able to do some things so much more quickly and efficiently that it wont be worthwhile to pay humans to do them anymore. This will include things like sensing, moving things around, scheduling, translating, and optimizing machinery. But the jury is still out over whether, in the long or short term, AI will lead to more jobs being lost or created.

One way to look at it is that AI, in theory, will lead to increased business growth and success. Often this will mean (hopefully) more customers. More customers mean more human problems that need to be solved from complicated customer service issues requiring a human response to the challenge of consistently creating innovative products and services that meet the changing needs of humans. These are tasks that humans will be needed for, for a long time yet!

The arrival of AI is described today as the dawn of the "fourth industrial revolution." The first industrial revolution saw political opposition and public unrest from people who feared that agricultural and manufacturing machinery would cause widespread unemployment and even the collapse of society. That fear is still alive today. On the other hand, others believe that technology will lead us into an era that has been described as fully automated luxury communism," where robots provide all our basic necessities at effectively no cost to us. Human beings are then free to spend their time on fulfilling and creative pursuits that give their lives value and meaning.

Both ideas present extreme outcomes, and were probably a long way yet from either. What is clear is that AI has the potential to relieve many of us of a lot of the mundane and repetitive elements of our work, so the best way to make sure we dont become redundant is to work in jobs where our value is elsewhere!

Smart machines will augment and assist us

When we're carrying out those more human, creative, or strategic tasks that won't be automated any time soon, we can expect robots and smart machines to be there to lend a hand. Often this will mean having analytics-capable tools on-hand to make sure our decisions are underpinned by solid data. For example, HR roles will, for a long time yet, still require a human touch to solve human-specific problems. But increasingly, AI is used in recruitment to provide initial screening of the thousands of applications that many large companies attract whenever they advertise a vacancy. You might not be happy to hand over the entire responsibility of picking who will work for you to a machine, but it can greatly improve efficiency by providing early indications of who might be the most suitable applicants.

Those kinds of tasks are carried out by AI software running on machines such as PCs and tablets that were all used to, but we will increasingly find ourselves working alongside machines in a very literal sense, too. Collaborative robots (Cobots) work on the floor with humans in Amazons worldwide network of warehouses as well as facilities such as Ericssons 5G smart factory, where assembly, packing, and dispatch of its devices is carried out autonomously by machines while security drones patrol the premises to deter intruders. Online supermarket Ocado uses robots to pick and pack 50,000 customer orders per hour, navigating miles of shelves across football-pitch-sized warehouses.

Our ability and willingness to get along with our new robot colleagues is likely to play a big part in determining how successful we are in the world of work in the near future. For businesses, the challenge will be to make sure humans and robots are both spending their time on the jobs they are best at.

AI will create new types of jobs

Once again, the impact of earlier industrial revolutions is a good source of predictions on how this one might turn out. Certainly, plowmen, weavers, and smiths lost their jobs during the widespread adoption of mechanization in the early 19th century. But as the first mechanized industries emerged, populations became urbanized, and quality of life improved for many people. This led to business and human enterprise evolving to provide services needed to keep everything running and to keep people fed, happy, and entertained.

The same will undoubtedly be true of the AI revolution. Business and society are going through a period of adjustment as we come to understand the power of smart machines and automation, and the human skills needed to lead the way on this are in high demand. More roles are likely to appear involving the ability to identify areas where AI and automation resources will be most effective, for example. Another very valuable human skill right now is the ability to create buy-in effectively developing trust among human workforces and bosses towards AI and smart machines because trust is essential for AI to work!

The key thing to remember here is that any job that can easily be automated is likely to be! Businesses need to ensure that their own workforce is in-tune with this forecasted future and equipped with the human skills that arent likely to be automated any time soon. These include emotional intelligence (empathy), creativity (AI can write songs but is unlikely to come up with anything that we will consider great music any time soon), and complex decision-making.

One way to think about the impact it will have, is to consider how the invention of "non-smart" machines (anything that can't be considered to be "artificially intelligent") impacted industry and commerce from the first industrial revolution to the dawn of the AI era, around ten years ago. During this time, machines took on much of the heavy lifting of our manual workload. Now we have "smart" machines; increasingly, they will take on the burden of managing our cerebral workload too anything that requires thought, learning, or decision making!

Read more about AI, technology, and other future trends in my new book, Business Trends in Practice: The 25+ Trends That are Redefining Organizations.

Read this article:
3 Ways That Artificial Intelligence (AI) Will Change Your Job Forever - Forbes