Daily Archives: May 27, 2022

(Aleutie/Shutterstock)

With inflation in the United States at a 40-year high and unemployment near a 50-year low, these are tough times to attract and retain employees in just about every sector. When you add the growing demand for talent in high tech sectors like big data and AI, you get a job market thats great for these workers, but tough for companies.

Whatever you call it the Great Resignation, the Great Reshuffle, or the Global Talent Shortage theres no denying that employers are under the gun when it comes to keeping skilled workers. Companies are scrambling to fill open positions in data and AI, let alone creating new ones to handle additional data and AI projects. This is causing employers to take drastic measures to keep up with the Joneses.

Here are three ways companies are dealing with the talent crunch:

This is probably the most obvious solution to attracting and retaining AI and analytics staffbut also the most painful for companies. With inflation currently at 8.3%, employees have a great excuse to seek higher pay, even if it results in higher costs and more inflation down the road. And with so much churn in the labor marketnearly 48 million people quit their jobs in 2021the conditions are perfect for them to get it.

Some tech firms are taking drastic measures. Microsoft for example is doubling its budget for employee salary increases, according to an article in Bloomberg. With the starting salary for a new engineer estimated to be around $160,000 per year, that is no small chunk of change for the second largest American company by market capitalization.

Inflation is expected to drive tech salaries up (Creativa Images/Shutterstock)

The move will help Microsoft to keep up with other tech giants eager to poach talent, including Amazon, which recently announced its doubling the maximum base salary for employees from $160,000 to $350,00 per year. That will certainly help to attract people who are looking for new jobs, which according to a recent survey accounted for 44% of all workers.

Companies will pay significant sums for coveted positions. According to a 2021 survey Hired conducted for its 2022 State of Software Engineers study, NLP engineers made an average of about $160,000 per year, machine learning engineers earned about $158,000, and data engineers grossed about $156,000.

The good news (for employers) is that salaries for these positions were flat relative to 2020. The biggest increase? Security engineers, which saw a 7.6% bump in salary to about $165,500.

2021 salaries may have been flat because they are a lagging indicator, according to the 2022 Dice Tech Salary Report. These high-growth and high-value occupations may begin to see an uptick in early 2022 and throughout the year, it suggested.

However, the odds of a recession have grown in recent weeks, as inflation takes a toll on consumer spending. That has led to speculation that the hiring binge will begin to slow. A spokesperson for Meta (parent company of Facebook) told CNBC earlier this week that the company was slowing its growth in hiring.

A recession would be painful for a lot of people and firms, but it likely would cool demand for tech talent.

Not too long ago, amenities like ping pong tables, bean bag chairs, and on-site chefs were enough to lure the best and brightest to tech startups. These days, folks are looking for something a little bit different, with a flexible working arrangement being near the top.

Postings for remote jobs on LinkedIn are getting a significantly higher response rate than jobs in specific locations (Source: LinkedIn)

Interest in jobs that allow workers to work from home is quite high. According to a post to the LinkedIn Talent Blog last month, remote jobs accounted for 20% of all job postings on LinkedIn, but accounted for 50% of all applications.

The message was crystal clear to Greg Lewis, the blogs author: As many companies seem eager to return workers to the office, candidates are sending a strong message that many of them would prefer to work remotely.

Data from a recent Pew Research study bears this out. Since 2020, the reason that people work from home has changed, the group says. During the early days of the pandemic, working from home was a matter of survival for the company, but not anymore.

Today, more workers say they are doing this by choice rather than necessity, Pew writes. Among those who have a workplace outside of their home, 61% now say they are choosing not to go into their workplace, while 38% say theyre working from home because their workplace is closed or unavailable to them.

For employers looking to satisfy a fickle workforce, allowing employees to work from home at least a few days a week could help keep them on the payrollfor at least a little longer.

Its long been observed that as technology improve, it displaces human workers. Weve seen this play out many times, including with the armies of clerks 100 years ago who manually tracked company spending on paper, only to be replaced with those Hollerith Tabulating Machines.

Africas demographics make it a promising location for BPO and IT outsourcing (monaliza0024/Shutterstock)

Fast forward to 2020 and the worst viral pandemic in decades, and automation is continuing to take over potentially dangerous jobs. For instance, toll booth operators for the Carquinez Bridge in the San Francisco Bay Area were replaced with FastTrak tags, displacing hundreds of workers, according to this story in Time.

In the world of analytics, the rise of self-service tools and techniques is helping to democratize data, but could it also make a dent in the hiring shortfall? According to Dice, which tracked a 11.5% increase in data analyst salaries from 2020 to 2021 (to about $85,000), the answer is yes.

For instance, numerous data-analytics apps allow employees of all backgrounds to crunch their organizations databases for key insights, Dice wrote in its recent salary survey. While these tools wont replace a highly specialized technologist, theyre a good way to streamline other employees workflows. With tech unemployment low and hiring managers having difficulty finding key talent, some organizations may be holding off on hiring some roles and relying on stopgap measures (and tools) instead.

Outsourcing also remains a possible tool to help companies through the Great Reshuffle. While its not really possible to outsource high-valued, strategic positions like data scientists, relying on business process outsourcing (BPO) providers to fill in for other positions can help companies free up resources and personnel to direct to the problem areas, which may include data and AI.

David Rickard of the Everest Group, a respected provider of insight for the global BPO industry, says that while countries like India have a lot to offer now, there are some other locales that should be on your radar, including Africa.

We talk about doubling down in India for the next three to five years in terms of looking for the talent, because theyve got the talent now, Rickard tells Datanami in a recent interveiw. But boldly go where no one has gone before and actually consider Africa as a long-term potential solution as it matures more and as people are coming into the workforce who are educated from an IT perspective.

Africa has a lot of things going for it, Rickard says. First and foremost, while the pipeline for workers entering the workforce in the future is shrinking in many developed countries, its actually getting bigger in Africa. If you look at the population in that 10 to 14 age range and the 15 to 19 age range, were talking about over a billion people coming into the workforce over the next few years, he says.

Everest Group ranks the countries across various criteria, including infrastructure, safety, security, economics, digital readiness, and quality of life, Rickard says. But then also we assess, whats the standard of English?

Tech giants are already investing in the leading countries. For example, Microsoft is investing in Rwanda, Rickard says, and Google is also making investments. In addition to Rwandan, other East African countries on Everests list include Kenya, Mauritius and Uganda. In West Africa, Ghana and Nigeria are good sources of workers from a BPO perspective, while in North Africa, its Egypt, Morocco, and Tunisia. South Africa also makes the list.

Rickard specializes in call center work, which is slightly different than IT work. But both require good education and English proficiency, so theres some possibility that Africa could play a bigger role in data work in the future.

Related Items:

Hiring, Pay for Data Science and Analytics Pros Picks Up Steam

In Search of Data Science Talent with Dr. Kirk Borne

Data Salaries Get a COVID Bump

Follow this link:

Three Ways Companies Can Cope with the AI and Analytics Talent Crunch - Datanami

In the opening remarks, Jiunn-Shiow Lin, the director of the IDB, summarized the AI development and strategies in Taiwan. Since 2019, to assist local companies in obtaining more business opportunities, the government-funded project, "AI Application Service Development Environment Promotion Program", has been exploring merging issues and trends for global AI applications.

The co-founder of the Centre for AI & Climate, Mr. Peter Clutton-Brock, portrayed a general picture of using AI to tackle climate change, from emerging technologies to possible solutions. Most importantly, he provided six potential fields in which AI can be the solution to the climate crisis. Followed by Dr. Vu Thuy Linh, the research fellow of AI for Operations Management Research Center, carried out an AI sensor system to reduce carbon emissions in smart buildings.

In terms of the agricultural production, Dr. I-Chun Chang, the general secretary of the Taiwan Smart Aquaculture Glow Association, shared their experience in promoting and implementing intelligent and automated modern production in Taiwanese aquaculture; Alan Yu, the founder of ID Water Technology Co., provided an AI-aid solution in shrimp farming which can boost the economic value compared with the traditional shrimp management methods.

When it comes to solving climatic problems, AI has been proven to be an accurate, fast, and reliable method to mitigate the effects of climate change on the economy, industry, and society. This webinar provided a viewpoint and hands-on experience in how companies can use AI to solve climatic problems across the world and strengthen the resiliency of businesses and societies.

For more information about this event, please visit: https://www.ai-hub.online/.

SOURCE AIHUB

Here is the original post:

Use AI in Tackling Climate Change: Experience Sharing from Taiwan and the World - PR Newswire

For a future fight against a near-peer military, U.S. special operators say they need smart, networked sensors and drones that can work together in contested environments with little human supervision. But as collaborative autonomy comes within technical reach, just how independent should these things get?

We are going to use a lot of sensors, whether they're unmanned aerial systems, unmanned ground systems, unmanned maritime systems, unintended sensors, all working together, and what our goal is to have those working together collaboratively and autonomously, SOCOMs top acquisition executive, James Smith, said at NDIAs SOFIC conference in Tampa, Florida, last week.

SOCOM has a specific line of effort where we're focused on what we're calling collaborative autonomy, said David Breede, who runs a program executive office at SOCOM. That line of effort is concerned with such questions as How do I get an unattended ground sensor talking to an unmanned aircraft and having an unmanned aircraft react based on the information that it got from that unmanned ground sensor? Not only collaboration across technologies and capabilities, but collaboration across program offices, right? Breede said. In other words, special operations forces need sensors on the ground, in the air, and in space constantly working together to autonomously detect changes and sound the alert.

But SOCOM doesnt just want networks of sensors to collaborate better. They also want the underlying autonomy software to work the same on everything from a 3-D printer to a $10,000 drone.

We have a goal of what we're calling portable autonomy, so being able to port software, virtual algorithms across different classes, of small drones. We would have an autonomy developer actually have their software algorithms on a payload and then integrate that onto a third-party platform and demonstrate the ability to control that platform without talking to that third-party platform provider, Breede said.

Among the obstacles: battlefield radio communications are expected to become much more difficult. SOCOMs Col. Paul Weizer said the command is trying to untether itself from radio.

So how do I operate completely in an untethered way, whether it's with unattended ground sensors or whether it's unmanned vehicles or otherwise?

Part of the answer is to put more information and computing power on the battlefield instead of counting on being able to reach back for it. The military and SOCOM in particular have been trying to bring cloud capabilities much closer to the battlefield, an effort exemplified by the Armys XVIII Airborne Corps and Amazon Web Services to create a tactical cloud environment.

That will also make battlefield decision-making much easier, said Quentin Donnellan, the president of the Space and Defense division of AI company Hypergiant, which is working with AWS and the Army on the effort.

If I turn on my radios, people are going to know where I'm at. So I don't want to turn on my radios, right? Donnellan said. So the idea for these use-cases is How do we deploy AI and machine learning out tactically where I can make those decisions in a communications-denied environment. If I've got the tools that allow me to, like, leverage AI to put it out to the edge, I should be able to do my job even if my cloud connection is denied.

One example of tactical cloud use is integrating radar sensor data for air defense in the fieldcloser to the threatrather than receiving an alert from a headquarters. That's kind of a really tactical and specific example of, Hey, if you deploy AI out there, [you could] potentially leverage weather or ground-based radar to be able to do things like object detection and classification, but not not relying on the connectivity back down, Donnellan said.

Shield AI co-founder Brandon Tseng said his companyknown for drones that navigate without GPSis working with SOCOM on portable autonomy to operate ever-larger drone swarms. Since 2018, Shield AI has been developing a software-based autonomy product called Hivemind for drone piloting; theyre integrating it ontoV-BAT drones to develop swarming and maritime domain awareness capabilities.

The company is working closely with the U.S. military to figure out how to penetrate enemy air defenses with drone swarms, he said. Something that we're super excited about is operationalizing swarms of three V-BAT aircraft in 2023, four craft in [20]24, eight aircraft in [20]25 and 16 aircraft in [20]26 that are working as a highly intelligent team together. I think it's adjacent to where SOCOM is and it definitely plays into their interests. But we're also integrating it on fighter jets and we expect to have it running on an F-16 later this year.

But the technology aspect of portable, collaborative autonomy isnt actually the hardest part of the challenge; the larger policy and ethics questions are.

Take the Switchblade from AeroVironment, the small kamikaze drones that have helped the Ukrainian military push back Russian forces. The drone sends video directly to an operator nearby without having to travel long distances over radio.

Brett Hush, vice president of tactical mission systems at AeroVironment, said his company is experimenting with artificial intelligence for automatic target recognition. Those capabilities are in development. We've demonstrated with the DOD our ability to do that to identify like 32 tanks and potentially strike them with no need for communication with an operator.

Now, fielding that capability is where we're gonna cross, you know, policy, he said. Today, everything that's done with our loitering missiles, there's a man on the loop. Once we go to field in the [automatic target recognition] and with more autonomy, we've got to really, as a country, think through where would that be allowed and not allowed.

View original post here:

'Collaborative, Portable Autonomy' Is the Future of AI for Special Operations - Defense One

When a new coronavirus emerged from nature in 2019, it changed the world. But COVID-19 wont be the last disease to jump across from the shrinking wild. Just this weekend, it was announced that Australia, is no longer an onlooker, as Canada, the US and European countries scramble to contain monkeypox, a less dangerous relative of the feared smallpox virus we were able to eradicate at great cost.

As we push nature to the fringes, we make the world less safe for both humans and animals. Thats because environmental destruction forces animals carrying viruses closer to us, or us to them. And when an infectious disease like COVID does jump across, it can easily pose a global health threat given our deeply interconnected world, the ease of travel and our dense and growing cities.

We can no longer ignore that humans are part of the environment, not separate to it. Our health is inextricably linked to the health of animals and the environment. This will not be the last pandemic.

To be better prepared for the next spillover of viruses from animals, we must focus on the connections between human, environmental and animal health. This is known as the One Health approach, endorsed by the World Health Organization and many others.

We believe artificial intelligence can help us better understand this web of connection, and teach us how to keep life in balance.

Fully 60% of all infectious diseases affecting humans are zoonoses, meaning they came from animals. That includes the lethal Ebola virus, which came from primates, swine flu, from pigs, and the novel coronavirus, most likely from bats. Its also possible for humans to give animals our diseases, with recent research suggesting transmission of COVID-19 from humans to cats as well as deer.

Early warning of new zoonoses is vital, if we are to be able to tackle viral spillover before it becomes a pandemic. Pandemics such as swine flu (influenza H1N1) and COVID-19 have shown us the enormous potential of AI-enabled prediction and disease surveillance. In the case of monkeypox, the virus has already been circulating in African countries, but has now made the leap internationally.

Read more: On the trail of the origins of Covid-19

What does this look like? Think of collecting and analysing real-time data on infection rates. In fact, AI was used to first flag the novel coronavirus as it was becoming a pandemic, with work done by AI company Bluedot and HealthMap at Boston Childrens Hospital.

How? By tracking vast flows of data in ways humans simply cannot do. Healthmap, for instance, uses natural language processing and machine learning to analyse data from government reports, social media, news sites, and other online sources to track the global spread of outbreaks.

We can also use AI to mine social media data to understand where and when the next COVID surge will occur. Other researchers are using AI to examine the genomic sequences of viruses infecting animals in order to predict whether they could potentially jump from their animal hosts into humans.

As climate change alters the earths systems, it is also changing the ways disease spreads and their distributions. Here, too, AI can be put to use in new surveillance methods.

There are clear links between our destruction of the environment and the emergence of new infectious diseases and zoonotic spillovers. That means protecting and conserving nature also helps our health. By keeping ecosystems healthy and intact, we can prevent future disease outbreaks.

In conservation, too, AI can help. For instance, Wildbook uses computer-vision algorithms to detect individual animals in images, and track them over time. This allows researchers to produce better estimates of population sizes.

Trashing the environment by deforestation or illegal mining can also be spotted by AI, such as through the Trends.Earth project, which monitors satellite imagery and earth observation data for signs of unwelcome change.

Citizen scientists can pitch in as well by helping train machine learning algorithms to get better at identifying endangered plants and animals on platforms like Zooniverse.

Researchers are beginning to consider the ethics of AI research on animals. If AI is used carelessly, we could actually see worse outcomes for domestic and wild animal species, for example, animal tracking data can be prone to errors if not double-checked by humans on the ground, or even hacked by poachers.

AI is ethically blind. Unless we take steps to embed values into this software, we could end up with a machine which replicates existing biases. For instance, if there are existing inequalities in human access to water resources, these could easily be recreated in AI tools which would maintain this unfairness. Thats why organisations such as the AINowInstitute are focusing on bias and environmental justice in AI.

In 2019, the EU released ethical guidelines for trustworthy AI. The goal was to ensure AI tools are transparent and prioritise human agency and environmental health.

Read more: How to prevent mass extinction in the ocean using AI, robots and 3D printers

AI tools have real potential to help us tackle the next pandemic by keeping tabs on viruses and helping us keep nature intact. But for this to happen, we will have to widen AI outwards, away from the human-centredness of most AI tools, towards embracing the fullness of the environment we live in and share with other species.

We should do this while embedding our AI tools with principles of transparency, equity and protection of rights for all.

Read this article:

AI could help us spot viruses like monkeypox before they cross over and help conserve nature - The Conversation

Omar Sultan Al Olama, the United Arab Emirates minister of artificial intelligence, yesterday told an audience at the World Economic forum in Davos that its his belief that people who commit serious crimes in the metaverse should be punished with real-world criminal consequences.

Per an article by CNBCs Sam Shead, the minister views this as a necessary measure to protect peoples mental health:

If I send you a text on WhatsApp, its text right? It might terrorize you but to a certain degree it will not create the memories that you will have PTSD (post-traumatic stress disorder) from it.

But if I come into the metaverse and its a realistic world that were talking about in the future and I actually murder you, and you see it it actually takes you to a certain extreme where you need to enforce aggressively across the world because everyone agrees that certain things are unacceptable.

Tell me you dont understand how post-traumatic stress disorder (PTSD) works without telling me you dont understand how PTSD works.

Subscribe now for a weekly recap of our favorite AI stories

Upfront: There is no medical threshold by which PTSD occurs. Clinical diagnosis involves observation and interviews with a medical professional.

Anecdotally speaking, PTSD isnt necessarily triggered in the manner which Al Olama indicated. I was diagnosed with PTSD while on active duty military service after learning about the death of someone Id mentored. Other peoples diagnosis have come after entirely different experiences.

Jennifer Kobelt, a survivor of the NXIVM cult, told investigators and documentarians that her PTSD was triggered after being subjected to a horrific experiment in which she was exposed to graphic violence from Hollywood cinema and a real-world snuff film.

Deeper: You cant murder an avatar. At least not in the legitimate legal sense. Its a stupid idea that doesnt deserve much attention, but lets just lay it bare real quick so we can move.

Lets say, 10 years from now, youre wandering around in Metas version of the metaverse. Youre probably wearing a VR headset, and maybe the techs advanced to the point where the visual and audio fidelity are nearly indistinguishable from reality.

All of the sudden, someone pushes the buttons on their control pad to cause their avatar to leap out of a digital bush and then they push the buttons on their control pad that cause them to stab your avatar.

Your avatar bleeds out and dies. You have to witness the knife going in! Oh! The horror!

But wait, lets rewind for a second. How did the knife get there? Who programmed the leaping out of the bush animation? Are there more kill moves? Whats the combo for a silent takedown?

Whoops. Im getting ahead of myself. I forgot, were not talking about a video game. Were talking about murder most foul, in the metaverse.

Im not sure what the UAEs minister of AI knows about the field that the rest of us dont, but in this particular version of reality, theres no basis for this fantasy.

Rock bottom: You may as well pass a law against murdering people in video games. And that means all of you people who play Call of Duty are screwed some of you have more kills than old age.

The point is that, no matter how traumatizing it might be to see yourself murdered in first person, its not like Zuckerbergs planning on making that a feature.

Maybe Al Olamas thinking the metaverse is going to be a splintered internet experience like web, where dark corners of the platform could be host to anything.

But, at least for now, the companies such as Meta, Nvidia, Microsoft, Google, and Epic that are investing billions of dollars into creating bespoke experiences probably arent going to put together a team of designers focused on adding PTSD-inducing gore to their production models.

Sure, a hacker could hack some violence onto a server or find an exploit that shows violence. And its possible some sort of underground mod scene could develop over time.

But seriously. The idea that somehow, youll be casually shopping in the Nike section of Metas billions-of-dollars and counting metaverse and suddenly a digital Jack the Ripper is going to appear in front of you in a rabid frenzy is just plain silly.

If you can murder people in the metaverse, itll be a feature that people log in specifically to experience. For the same reason so many of us play Dead By Daylight, Resident Evil, and Call of Duty, or watch R-rated horror movies, theres plenty of people whod enjoy a good old-fashioned fake-murdering in a VR world.

Quick take: Everything about the idea of criminalizing digitized violence in virtual reality is dumb. This kind of blathering rhetoric just demonstrates how far detached from reality some technologists can be. Nobodys worried about logging onto a VR version of Facebook and being murdered in their headset.

There are plenty of real ethical concerns that the minister of AI for the sixth richest country in the world could spend their time on.

Read more:

The UAE's AI minister wants 'murder' in the metaverse to be a real crime - The Next Web

MANCHESTER, United Kingdom, May 26, 2022 (GLOBE NEWSWIRE) -- Most businesses (90%) now use or plan to use Artificial Intelligence (AI) and even more (98%) have, or intend to, implement a data strategy. Yet new findings suggest that while the success of each is reliant on the other only one in three (35%) businesses with a data strategy say it includes provisions for Artificial Intelligence (AI). Decision Intelligence company Peak, released its State of AI, 2022 report today, highlighting that many businesses are making investments in both data and AI without connecting the two.

AI is inherently a data technology, and must function as part of an overarching data strategy, yet the majority of respondents in this survey are thinking about the two separately, said Richard Potter, co-founder and CEO of Peak.

The amount of data businesses produce is increasing exponentially and current predictions suggest businesses globally will create, capture, copy and consume 97 zettabytes of data this year alone.1 Implementing a data strategy is essential for defining how to collect, store and put that data to use.

Peaks State of AI report suggests that effectively leveraging their data is a priority for many businesses. One in two (52%) organizations with 100 or more employees now have a Chief Data Officer, and 86% have invested in a data lake or warehouse. Still, two-thirds (65%) of those with a data strategy have not made provisions for AI. Instead, data strategies are most likely to be focused on centralizing data within the business (55%), although establishing measurable goals for the use of data (54%) and managing security risks and compliance (53%) are similarly important goals.

The report also uncovered inconsistencies among leadership teams on data strategy. While 89% of CEOs said their company had a data strategy, only 63% of wider c-suite executives agreed.

This disconnect extends to perceptions of AI as well. Of the respondents in businesses that have or are working towards AI, 55% of CEOs thought the technology was being implemented throughout the business, compared to just 42% of other c-suite leaders.

Our research also reveals that within many companies there is a lack of clarity around the overall AI strategy, even at the top levels of management, said Richard Potter. If businesses want to successfully implement AI and, critically, drive value with this technology, we need open discussion within a business to ensure everyone is aligned to and understands the vision.

The report also provides a look at the regional differences in attitudes and progress being made in digital transformation, data maturity and AI adoption, with India more advanced than both the US and UK.

Methodology The report is based on a survey of 775 decision makers (senior managers and above) from the US, UK or India. Respondents were all from companies with 100 or more staff. Research was conducted for Peak by Opinium, in partnership with the Center for Economics and Business Research (Cebr). The survey ran from 19-25 April, 2022.

About Peak Jointly founded in Manchester and Jaipur in 2015 by Richard Potter, David Leitch and Atul Sharma, Peak is on a mission to change the way the world works.

A pioneer of the Decision Intelligence category, its platform enables customers to apply AI to the commercial decision making process. Peak features three products Dock, Factory, and Work that can be leveraged by businesses at every stage of their AI journey to build apps that deliver against specific business needs. With features to support both technical and line-of-business users, Peak makes these apps widely accessible to everyone within a business, simplifying and accelerating adoption of AI. It is used by leading brands including Nike, ASOS, PepsiCo, KFC and Sika.

The company has grown significantly over the last three years and, in August 2021, Peak announced a $75m Series C funding round led by SoftBanks Vision Fund II. The same year, it received a Best Companies 3-star accreditation, which recognises extraordinary levels of employee engagement and was ranked by The Sunday Times as one of the Best 100 Companies to Work For in 2020 and 2021.

1 Statista, 2022. Volume of data/information created, captured, copied, and consumed worldwide from 2010 to 2025 (in zettabytes). Statista.

Contact:

Mica Hahnpeak@praytellagency.com510.206.3932

More here:

Commercial AI Adoption is Picking Up Pace But Only A Third of Businesses Recognise that Data Strategy Facilitates AI - GlobeNewswire

On May 12, the Equal Employment Opportunity Commission and the U.S. Department of Justice issued guidance to caution employers about using artificial intelligence (AI) and software tools to make employment decisions. The guidance, titled The Americans with Disabilities Act and the Use of Software, Algorithms, and Artificial Intelligence to Assess Job Applicants and Employees, warns that using these tools without safeguards could result in an Americans with Disabilities Act violation.

The ADA requires that people with disabilities have full access to public and private services and facilities. It also bans employers from discriminating on the basis of disabilities, and requires that they provide reasonable accommodations to employees with disabilities to enable them to do a job.

Companies increasingly rely on software and AI to monitor employees locations, productivity, and performance, such as surveillance monitoring that tracks employees time on task. Employers may then use this information to make pay, disciplinary, and termination decisions. The guidance warns employers that doing so may result in discrimination against employees with disabilities.

The EEOC advises that employers should not be using algorithms, software or AI to make employment decisions without giving employees a chance to request reasonable accommodations. Heres why: Software and AI are designed to provide information based on preset specifications of the average or ideal worker. However, employees with disabilities may not work under typical conditions if they are utilizing an accommodation.

To use these tools without running afoul of the law, employers should train their HR personnel, managers and supervisors to recognize and respond to requests for accommodations from employees, even when the request is informal. This could include an employees requests to take a test in an alternative format, to be assessed in another way, or to be allowed to spend extra time off task due to a medical condition, for example. Employers should also ensure they use software that has been tested by users with disabilities. Other tips for staying in compliance include providing clear instructions to employees for requesting accommodations and avoiding pre-employment screening for traits that may reveal disabilities.

Employers would do well to remember that employees are humans, and sometimes decisions about humans need to be made by humans, not by computers. Metrics tracked by computers should be just one piece of the puzzle when evaluating employees performance.

Read more here:

Algorithms, AI, and the ADA | McAfee & Taft - JDSupra - JD Supra

As the demand for AI-powered apps grows, startups developing dedicated chips to accelerate AI workloads on-premises are reaping the benefits. A recent ZDNet piece reaffirms that the AI edge chip market is booming, fueled by staggering venture capital financing in the hundreds of millions of dollars. EdgeQ, Kneron and Hailo are among the dozens of upstarts vying for customers, the last of which nabbed $136 million in October as it doubles down on new opportunities.

Another company competing in the increasingly saturated segment is SiMa.ai, which is developing a system-on-chip platform for AI applications particularly computer vision applications. After emerging from stealth in 2019, SiMa.ai began demoing an accelerator chipset that combines traditional compute IP from Arm with a custom machine learning accelerator and dedicated vision accelerator, linked via a proprietary interconnect,

To lay the groundwork for future growth, SiMa.ai today closed a $30 million additional investment from Fidelity Management & Research Company, with participation from Lip-Bu Tan (whos joining the board) and previous investors, concluding the startups Series B. It brings SiMa.ais total capital raised to $150 million.

The funding will be used to accelerate scaling of the engineering and business teams globally, and to continue investing in both hardware and software innovation, founder and CEO Krishna Rangasayee told TechCrunch in an email interview. The appointment of Lip-Bu Tan as the newest member of SiMa.ais board of directors is a strategic milestone for the company. He has a deep history of investing in deep tech startups that have gone on to disrupt industries across AI, data, semiconductors, among others.

Rangasayee spent most of his career in the semiconductor industry at Xilinx, where he was GM of the companys overall business. An engineer by trade Rangasayee was the COO at Groq and once headed product planning at Altera, which Intel acquired in 2015 he says that he was motivated to start SiMa.ai by the gap he saw in the machine learning market for edge devices.

I founded SiMa.ai with two questions: What are the biggest challenges in scaling machine learning to the embedded edge? and How can we help?, Rangasaye said. By listening to dozens of industry-leading customers in the machine learning trenches, SiMa.ai developed a deep understanding of their problems and needs like getting the benefits of machine learning without a steep learning curve, preserving legacy applications along with future proof ML implementations, and working with a high-performance, low-power solution in a user-friendly environment.

SiMa.ai aims to work with companies developing driverless cars, robots, medical devices, drones and more. The company claims to have completed several customer engagements and last year announced the opening of a design center in Bengaluru, India, as well as a collaboration with the University of Tbingen to identify AI hardware and software solutions for ultra-low energy consumption.

As over-100-employee SiMa.ai works to productize its first-generation chip, work is underway on the second-generation architecture, Rangasayee said.

SiMa.ais software and hardware platform can be used to enable scaling machine learning to [a range of] embedded edge applications. Even though these applications will use many diverse computer vision pipelines with a variety of machine learning models, SiMa.ais software and hardware platform has the flexibility to be used to address these, Rangasayee added. SiMa.ais platform addresses any computer vision application using any model, any framework, any sensor, any resolution [We as a company have] seized the opportunity to disrupt the burgeoning edge computing space in pursuit of displacing decades-old technology and legacy incumbents.

SiMa.ais challenges remain mass manufacturing its chips affordably and beating back the many rivals in the edge AI computing space. (The companys says that it plans to ship mass-produced production volumes of its first chip sometime this year.) But the startup stands to profit handsomely if it can capture even a sliver of the sector. Edge computing is forecast to be a $6.72 billion market by 2022, accordingto Markets and Markets. Its growth will coincide with that of the deep learning chipset market, whichsome analysts predict will reach $66.3 billion by 2025.

Machine learning has had a profound impact on the cloud and mobile markets over the past decade and the next battleground is the multi-trillion-dollar embedded edge market, Tan said in a statement. SiMa.ai has created a software-centric, purpose-built platform that exclusively targets this large market opportunity. SiMa.ais unique architecture, market understanding and world-class team has put them in a leadership position.

Read the original here:

AI chip startup SiMa.ai bags another $30M ahead of growth - TechCrunch

A century ago, English mathematician Lewis Fry Richardson proposed a startling idea for that time: constructing a systematic process based on math for predicting the weather. In his 1922 book, Weather Prediction By Numerical Process, Richardson tried to write an equation that he could use to solve the dynamics of the atmosphere based on hand calculations.

It didnt work because not enough was known about the science of the atmosphere at that time. Perhaps some day in the dim future it will be possible to advance the computations faster than the weather advances and at a cost less than the saving to mankind due to the information gained. But that is a dream, Richardson concluded.

A century later, modern weather forecasts are based on the kind of complex computations that Richardson imagined and theyve become more accurate than anything he envisioned. Especially in recent decades, steady progress in research, data and computing has enabled a quiet revolution of numerical weather prediction.

For example, a forecast of heavy rainfall two days in advance is now as good as a same-day forecast was in the mid-1990s. Errors in the predicted tracks of hurricanes have been cut in half in the last 30 years.

There still are major challenges. Thunderstorms that produce tornadoes, large hail or heavy rain remain difficult to predict. And then theres chaos, often described as the butterfly effect the fact that small changes in complex processes make weather less predictable. Chaos limits our ability to make precise forecasts beyond about 10 days.

As in many other scientific fields, the proliferation of tools like artificial intelligence and machine learning holds great promise for weather prediction. We have seen some of whats possible in our research on applying machine learning to forecasts of high-impact weather. But we also believe that while these tools open up new possibilities for better forecasts, many parts of the job are handled more skillfully by experienced people.

Today, weather forecasters primary tools are numerical weather prediction models. These models use observations of the current state of the atmosphere from sources such as weather stations, weather balloons and satellites, and solve equations that govern the motion of air.

These models are outstanding at predicting most weather systems, but the smaller a weather event is, the more difficult it is to predict. As an example, think of a thunderstorm that dumps heavy rain on one side of town and nothing on the other side. Furthermore, experienced forecasters are remarkably good at synthesizing the huge amounts of weather information they have to consider each day, but their memories and bandwidth are not infinite.

Artificial intelligence and machine learning can help with some of these challenges. Forecasters are using these tools in several ways now, including making predictions of high-impact weather that the models cant provide.

In a project that started in 2017 and was reported in a 2021 paper, we focused on heavy rainfall. Of course, part of the problem is defining heavy: Two inches of rain in New Orleans may mean something very different than in Phoenix. We accounted for this by using observations of unusually large rain accumulations for each location across the country, along with a history of forecasts from a numerical weather prediction model.

We plugged that information into a machine learning method known as random forests, which uses many decision trees to split a mass of data and predict the likelihood of different outcomes. The result is a tool that forecasts the probability that rains heavy enough to generate flash flooding will occur.

We have since applied similar methods to forecasting of tornadoes, large hail and severe thunderstorm winds. Other research groups are developing similar tools. National Weather Service forecasters are using some of these tools to better assess the likelihood of hazardous weather on a given day.

Researchers also are embedding machine learning within numerical weather prediction models to speed up tasks that can be intensive to compute, such as predicting how water vapor gets converted to rain, snow or hail.

Its possible that machine learning models could eventually replace traditional numerical weather prediction models altogether. Instead of solving a set of complex physical equations as the models do, these systems instead would process thousands of past weather maps to learn how weather systems tend to behave. Then, using current weather data, they would make weather predictions based on what theyve learned from the past.

Some studies have shown that machine learning-based forecast systems can predict general weather patterns as well as numerical weather prediction models while using only a fraction of the computing power the models require. These new tools dont yet forecast the details of local weather that people care about, but with many researchers carefully testing them and inventing new methods, there is promise for the future.

There are also reasons for caution. Unlike numerical weather prediction models, forecast systems that use machine learning are not constrained by the physical laws that govern the atmosphere. So its possible that they could produce unrealistic results for example, forecasting temperature extremes beyond the bounds of nature. And it is unclear how they will perform during highly unusual or unprecedented weather phenomena.

And relying on AI tools can raise ethical concerns. For instance, locations with relatively few weather observations with which to train a machine learning system may not benefit from forecast improvements that are seen in other areas.

Another central question is how best to incorporate these new advances into forecasting. Finding the right balance between automated tools and the knowledge of expert human forecasters has long been a challenge in meteorology. Rapid technological advances will only make it more complicated.

Ideally, AI and machine learning will allow human forecasters to do their jobs more efficiently, spending less time on generating routine forecasts and more on communicating forecasts implications and impacts to the public or, for private forecasters, to their clients. We believe that careful collaboration between scientists, forecasters and forecast users is the best way to achieve these goals and build trust in machine-generated weather forecasts.

Follow this link:

AI and machine learning are improving weather forecasts, but they won't replace human experts - The Conversation Indonesia

World Weather Attribution

It sounds like a contradiction in terms, but disaster and disruption management is a thing. Disaster and disruption are precisely what ensues when catastrophic natural events occur, and unfortunately, the trajectory the world is on seems to be exacerbating the issue. In 2021 alone, the US experienced15+ weather/climate disaster eventswith damages exceeding $1 billion.

Previously, we have explored various aspects of the ways data science and machine learning intertwine with natural events -- fromweather predictiontothe impact of climate change on extreme phenomenaandmeasuring the impact of disaster relief. AiDash, however, is aiming at something different: helping utility and energy companies, as well as governments and cities, manage the impact of natural disasters, including storms and wildfires.

We connected with AiDash co-founder and CEO Abhishek Singh to learn more about its mission and approach, as well its newly releasedDisaster and Disruption Management System (DDMS).

Singh describes himself as a serial entrepreneur with multiple successful exits. Hailing from India, Singh founded one of the world's first mobile app development companies in 2005 and then an education tech company in 2011.

Following the merger of Singh's mobile tech company with a system integrator, the company was publicly listed, and Singh moved to the US. Eventually, he realized that power outages are a problem in the US, with the wildfires of 2017 were a turning point for him.

That, and the fact that satellite technology has been maturing -- with Singh marking 2018 as an inflection point for the technology -- led to founding AiDash in 2020.

AiDash notes that satellite technology has reached maturity as a viable tool. Over 1,000 satellites are launched every year, employing various electromagnetic bands, including multispectral bands and synthetic aperture radar (SAR) bands.

The company uses satellite data, combined with a multitude of other data, and builds products around predictive AI models to allow preparation and resource placement, evaluate damages to understand what restoration is needed and which sites are accessible and help plan the restoration itself.

AiDash uses a variety of data sources. Weather data, to be able to predict the course storms take and their intensity. Third-party or enterprise data, to know what assets need to be protected and what their locations are.

Also:The EU AI Act could help get to Trustworthy AI, according to the Mozilla Foundation

The company's primary client thus far has been utility companies. For them, a common scenario involves damages caused by falling trees or floods. Vegetation, in general, is a key factor in AiDash AI models but not the only one.

As Singh noted, AiDash has developed various AI models for specific use cases. Some of them include an encroachment model, an asset health model, a tree health model and an outage prediction model.

Those models have taken considerable expertise to develop. As Singh noted, in order to do that, AiDash is employing people such as agronomists and pipeline integrity experts.

"This is what differentiates a product from a technology solution. AI is good but not good enough if it's not domain-specific, so the domain becomes very important. We have this team in-house, and their knowledge has been used in building these products and, more importantly, identifying what variables are more important than others", said Singh.

To exemplify the application of domain knowledge, Singh referred to trees. As he explained, more than 50% of outages that happen during a storm are because of falling trees. Poles don't normally fall on their own -- generally, it's trees that fall on wires and snap them or cause poles to fall. Therefore, he added that understanding trees is more important than understanding the weather in this context.

"There are many weather companies. In fact, we partner with them -- we don't compete with them. We take their weather data, and we believe that the weather prediction model, which is also a complicated model, works. But then we supplement that with tree knowledge", said Singh.

In addition, AiDash uses data and models about the assets utilities manage. Things such as what parts may break when lightning strikes, or when devices were last serviced. This localized, domain-specific information is what makes predictions granular. How granular?

Also:Averting the food crisis and restoring environmental balance with data-driven regenerative agriculture

Supplementing data and AI models with domain-specific knowledge, in this case knowledge about trees, is what makes the difference for AiDash

"We know each and every tree in the network. We know each and every asset in the network. We know their maintenance history. We know the health of the tree. Now, we can make predictions when we supplement that with weather information and the storm's path in real-time. We don't make a prediction that Texas will see this much damage. We make a prediction that this street in this city will see this much damage," Singh said.

In addition to utilizing domain knowledge and a wide array of data, Singh also identified something else as key to AiDash's success: serving the right amount of information to the right people the right way. All the data live and feed the elaborate models under the hood and are only exposed when needed -- for example if required by regulation.

For the most part, what AiDash serves is solutions, not insights, as Singh put it. Users access DDMS via a mobile application and a web application. Mobile applications are meant to be used by people in the field, and they also serve to provide validation for the system's predictions. For the people doing the planning, a web dashboard is provided, which they can use to see the status in real-time.

Also:H2O.ai brings AI grandmaster-powered NLP to the enterprise

DDMS is the latest addition toAiDash's product suite, including the Intelligent Vegetation Management System, the Intelligent Sustainability Management System, the Asset Cockpit and Remote Monitoring & Inspection. DDMS is currently focused on storms and wildfires, with the goal being to extend it to other natural calamities like earthquakes and floods, Singh said.

The company's plans also include extending its customer base to public authorities. As Singh said, when data for a certain region are available, they can be used to deliver solutions to different entities. Some of these could also be given free of charge to government entities, especially in a disaster scenario, as AiDash does not incur an incremental cost.

AiDash is headquartered in California, with its 215 employees spread in offices in San Jose and Austin in Texas, Washington DC, London and India. The company also has clients worldwide and has been seeing significant growth. As Singh shared, the goal is to go public around 2025.

Link:

Managing disaster and disruption with AI, one tree at a time - ZDNet