Cloud Hosting

LAWRENCE If youve ever used a text-based artificial-intelligence image generator like Craiyon or DALL-E, you know with a few word prompts that the AI tools create images that are both realistic and completely synthesized.

The machine learning that powers such websites will scan millions of images on the internet, analyze them and assemble facets of them into fresh, but fake, images.

Now, University of Kansas researchers are working to use a similar machine-learning process to build new proteins designed to detect water pollutants. With a new three-year, $1.5 million grant from the National Science Foundations Molecular Foundations for Biotechnology program, a KU researcher will use machine learning to create deep-fake membrane beta-barrel proteins a class of naturally successful biosensors designed to detect polluting metal ions in water.

These beta barrels are super useful because they can bring things across membranes, said principal investigator Joanna Slusky, associate professor of molecular biosciences at KU. Barrels make good enzymes there are so many different things that barrels can do.

Previous research on the tube-like beta barrels has altered their binding properties for a variety of tasks. However, much of this work was arduous and completed by hand, usually resulting with minor variations of a limited number of scaffolds, or barrel structures.

In this case, were using machine learning to generate large numbers of barrels, Slusky said. But, how about if we can both generate barrels and have them be useful? We asked ourselves, What's a biotechnology application of barrels? Well, one would be metal sensors that could perhaps detect metal pollutants.

Slusky and her co-principal investigators, professors Rachel Kolodny and Margarita Osadchy of Haifa University in Israel (along with KU postdoctoral fellow Daniel Montezano), will develop a new machine-learning process that generates beta-barrels with scaffolds similar to those found in nature, but with different sequences.

Theres a website called This X Does Not Exist, Slusky said. If you go to that site, you see all these AI-generated things and people don't really exist. But a computer made an image, for instance, of a cat. But that's not really a cat a computer took a bunch of pictures of cats and said, OK, we can just sort of generate as many cat pictures as you want now, because we figured out what is a cat. We need to make something real so we see it more like generating a recipe.

"The question is, how to make computers generate a recipe for proteins.

Beta barrels are well-suited to advancement through machine learning because natural proteins are sort of a small blip in the number of possible sequences.

If a computer algorithm can learn the essence of what makes a protein a protein, Slusky said, it will avoid generating useless sequences.

Most sequences would never actually be proteins they wouldn't have a particular fold, she said. They would just kind of bond with themselves in weird, nonpredictable ways over and over again. To be a protein, you need a sequence that makes one shape. When people tried to make random sequences, or even somewhat directed sequences, they found that only a very, very small percentage of them might actually be a protein.

With machine learning creating new and viable sequences resulting in this common fold, Slusky and her colleagues hope to generate a beta-barrel especially well-suited to finding metal ions in water. This result of the work will be biosensors based on beta barrels that can identify pollutants like lead in waterways.

If we make them the right size, this molecule will be ideal to put some particular metal in, and you can have the right substituents so that it would bind that metal, Slusky said. Because it's in a membrane, it can give you some sort of conductance difference theres a difference between when it's bound and when it's not bound. If youre able to do that, you could sense for different metals, and different concentrations of those metals. There are a lot of big steps we want to accomplish, but Im hopeful and excited.

The work also will help train undergraduate researchers in Sluskys lab, as well as inform Sluskys teaching at KU as well as outreach to high-school science students.

Top right image:A University of Kansas researcher will use machine learning to create deep-fake membrane beta-barrel proteins a class of naturally successful biosensors designed to detect polluting metal ions in water. Credit:T. Chris Gamblin.

Bottom right image: Joanna Slusky, associate professor of molecular biosciences at KU.

View post:
'Deep fake' protein designed with artificial intelligence will target water pollutants | The University of Kansas - The University of Kansas

(Bloomberg) -- Staring closely at the screen, Jordan Wright deftly picks out a barely distinguishable shape with his mouse, bringing to life a stark blue outline from a blur of overexposed features.

Its a process similar to the automated tests that teach computers to distinguish humans from machines, by asking someone to identify traffic lights or stop signs in a picture known as a Captcha.

Related: Cisco Fires Workers for Racial Comments During Diversity Forum

Only in Wrights case, the shape turns out to be of a Tupolev Tu-160, a supersonic strategic heavy bomber, parked on a Russian base. The outline one of hundreds a day he picks out from satellite imagesis training an algorithm so a US intelligence agency can locate and identify Moscows firepower in an automated flash.

Its become a run-of-the-mill task for the 25-year-old, who describes himself as on the autism spectrum. Starting in the spring, Wright began working atEnabled Intelligence, a Virginia-based startup that works largely for US intelligence and other federal agencies. Foundedin 2020, itspecializes in labeling, training and testing the sensitive digital data on which artificial intelligence depends.

Related: FastChat: Bringing Diversity and Inclusion to the Data Center Space

Wright works at the Virginia-based startup that utilizes a workforce described as neurodiverse.

Peter Kant, chief executive officer of Enabled Intelligence, said he was inspired to start the company after reading about an Israeli program to recruit people with autism for cyber-intelligence work. Therepetitive,detailedwork of training artificial intelligence algorithms relies on pattern recognition, puzzle-solving and deep focus that is sometimes a particular strength of autistic workers, he said.

Enabled Intelligences main type ofwork, known as data annotation, is usually farmed out to technically skilled but far cheaper labor forces in countries including China, Kenya andMalaysia. Thats not an option for US government agencies whose data is sensitive or classified, Kant said, adding that morethan half hisworkforce of 25 areneurodiverse.

I can easily say this is the best opportunity I've got in my life, said Wright, who grew up with an infatuation for military aviation, dropped out of college and has since experienced long stints of unemployment in between poorly paid work. Most recently, he baggedfrozen groceries.

For decades, workers with developmental disabilities, especially autism, have faced discrimination and disproportionately high unemployment levels. A large shortfall in cybersecurity jobs, along with a new push for workplace acceptance and flexibility in part spurred by the Covid-19 pandemic has started to focus attention onthe abilities of people who think and work differently.

Enabled Intelligence has adjusted its work rules to accommodate its employees, ditching resumes and interviews for online assessmentsand staggering work hours for those who find it hard to get in early. It has built three new areas for classified material and hopes to secure government clearances for much of its neurodiverse workforce something the US intelligence community has sometimes struggled to accommodate in the past.Pay starts at $20 an hour,in line with industry standards, and the company provideshealth insurance, paid leave and a path for promotion. Enabled Intelligenceexpects to make revenues of $2 millionthis year and double thatnext year, along with doubling its workforce.

The US intelligence community has been slow to catch on to the opportunity, critics say. It falls short of the 12% federal target for workforce representation of persons with disabilities, according to the lateststatisticsout this month. Until this year, it has also regularlyfallen shortof the 2% federal target for persons with targeted disabilities, which include those with autism.

In other countries its old hat, said Teresa Thomas, program lead for neurodiverse talent enablement at MITRE, which operates federally funded research and development centers. She citeswell established programs in Denmark, Israel, the UK and Australia, where one state recently appointed a minister for autism.

Thomas has recently spearheaded a new neurodiverse federal workforce pilot to establish a template for the US government to hire and support autistic workers, but so far only one of the countrys 18 intelligence agencies, the National Geospatial-Intelligence Agency, known asNGA,has participated.Now the federal governmentscyberdefense agency, the Cybersecurity and Infrastructure Security Agency,intends to undertake a similar pilot.

Stephanie La Rue, chief of diversity, equity and inclusion for the Office of the Director of National Intelligence, told Bloomberg the US intelligence community needs to acknowledge that its not where we need to bewhen it comes to employing people with disabilities.

Its like turning the Titanic, said La Rue, adding that NGAs four-person pilot would be reviewed and shared with the wider intelligence community as a promising practice. Change is going to be incremental.

Research indicated that neurodiverse intelligence officers on the autism spectrum exhibit the ability to parse large data sets and identify patterns and trends at rates that far exceed folks who are not autistic and were less prone to cognitive bias, La Rue said.Yet securing a clearance to access classified information can still present an additional challenge, according to some observers.

Enabled Intelligence CEO Peter Kant, standing, was inspired to start the company after reading about an Israeli program to recruit people with autism for cyberintelligence work Photographer: Valerie Plesch/Bloomberg

If an office wall board at Enabled Intelligenceis any indication, experiencesvary. There, 18 anonymous handwritten notesanswer the question: What does neurodiversity mean to you?

Difficult. Trying. Its held me back a lot, says one in an uncertain script. Strength,answers a second in careful cursive. A third, in capital letters, declares: SUPERPOWERS.

Follow this link:
Neurodiversity Emerges as a Skill in Artificial Intelligence Work - Data Center Knowledge

AI's future is commercial.

In line with a rocky and uncertain economic climate, the pace of investments flowing into the red-hot artificial intelligence technology space has cooled somewhat this past year. Things are still red hot, however, and AI is seeing a lot of progress, mitigated by concerns over safety and responsibility. Interestingly, much of its development has moved out of labs and into commercial ventures.

These are the conclusions drawn by two leading venture capitalists in the tech space, Nathan Benaich of Air Street Capital and Ian Hogarth of Plural, outlined in their annual summary of the state of AI. The report covers all facets of AI, from developments with DeepMind to NVIDIAs rapidly expanding processing capabilities. There are also numerous implications for AI from a business perspective.

For starters, it turns out that 2021 was a banner year for the AI business sector, but then softened in 2022. In 2022, investment in startups using AI has slowed down along with the broader market. Private companies using AI are expected to raise 36% less money in 2022 versus the previous year, but are still on track to exceed the 2020 level. This is comparable with the investment in all startups and scaleups worldwide, they observe. In addition, they note, enterprise software is the most invested category globally, while robotics captures the largest share of VC investment into AI.

At the same time, there has been a softening, though less extreme, for investments in SaaS startups and scaleups using AI expected to reach $41.5 billion by the end of the year, down 33% from last year. This is still higher than in 2020 VC investment in AI SaaS startups and scaleups.

Significantly, the reports co-authors observe, there has also been a drying up of academic research in AI as multi-year project funding concludes, with much of the research now shifted to the commercial sector. That means more startups and scaleups on the horizon. Once considered untouchable, talent from Tier 1 AI labs is breaking loose and becoming entrepreneurial, Benaich and Hogarth state. Alums are working on AGI, AI safety, biotech, fintech, energy, dev tools and robotics.

They add that hiring freezes and the disbanding of AI labs precipitates the formation of many startups from giants including DeepMind and OpenAI. Even the large tech behemoths are seeing some loss of talent to startups. Meta, for example, is folding their centralized AI research group after letting it run free from product roadmap pressure for almost 10 years. In addition, all but one author of the landmark paper that introduced transformer-based neural networks have left Google to build their own startups in artificial general intelligence, conversational agents, AI first biotech and blockchain, they note. For example, they relate, AnthropC raised $580 million in 2022, Inflection raised $225 million, and co:here raised $125 million.

Worldwide Investment in Startups and Scaleups Using AI:

Benaich and Hogarth also looked at the prevalence of AI unicorns emerging across nations of the world. concluding the United States leads in these high-potential startups, followed by China and the United Kingdom. A total 292 AI unicorns emerged within the US in 2022, with a combined enterprise value of $4.6 trillion. Overall, they add, despite significant drop in investment in US based startups and scaleups using AI, they still account for more than half of the AI investment worldwide.

Also in 2022, the big tech companies continued to expand their AI clouds and form large partnerships with AI startups, Benaich and Hogarth state. The hyperscalers and challenger AI compute providers are tallying up major AI compute partnerships, notably Microsofts $1 billion investment into OpenAI. We expect more to come.

For the year ahead, Benaich and Hogarth predict more than $100 million will be invested in dedicated AI-alignment organizations in the next year as more people become aware of the risk we are facing by letting AI capabilities run ahead of safety. In addition, they predict that a major user-generated content side will negotiate a commercial settlement with a startup producing AI models (such as OpenAI) for training on their corpus of user generated content.

Original post:
Pace Of Artificial Intelligence Investments Slows, But AI Is Still Hotter Than Ever - Forbes

ARLINGTON, Va. U.S. military researchers will brief industry next month on an upcoming project to develop new kinds of artificial intelligence (AI) and machine autonomy for battle management and sensor fusion.

Officials of the U.S. Defense Advanced Research Projects Agency (DARPA) in Arlington, Va., issued a special notice (DARPA-SN-23-06) on Monday for the Artificial Intelligence Reinforcements (AIR) project.

The DARPA AIR initiative seeks to fill gaps in research on developing and deploying tactical autonomy capability in real-world military operations. Industry briefings will be from 8:30 a.m. to 5 p.m. on Monday 14 Nov. 2022 at Amentums Ballston Conference Center, 4121 Wilson Blvd., in Arlington, Va.

AIR will focus on previously avoided dimensions to enable tactical autonomy in integrated sensors, scalability to large engagements, adaptability to changing conditions, and the ability to learn predictive models that incorporate uncertain knowledge of adversary and self, as well as deceptive effects.

Related: Marines ask Sentient Vision for artificial intelligence (AI) and machine autonomy for unmanned reconnaissance

AIR will pair existing, maturing, and emerging algorithmic approaches with expert human feedback to evolve the cooperative autonomous behaviors rapidly that solve previously avoided challenges.

AIR will address two technical areas: creating fast and accurate models that capture uncertainty and automatically improve with more data; and developing AI-driven algorithmic approaches to real-time distributed autonomous tactical execution within uncertain, dynamic, and complex operational environments.

The AIR program also will develop ways to design, test, and implement future iterations of AIR software.

Related: Artificial intelligence (AI) to enable manned and unmanned vehicles adapt to unforeseen events like damage

Briefings will be at Amentums Ballston Conference Center on the second floor of 4121 Wilson Blvd. in Arlington, Va., on Monday 14 Nov. 2022 from 8:30 a.m. to 5:00 p.m. Check-in begins at 8 a.m.

Briefings will include information that is International Traffic in Arms (ITAR)-restricted, so attendance is limited to U.S. citizens or U.S. permanent residents representing U.S. companies. Briefings will be classified at the Collateral SECRET level and will require security clearances.

Those interested in attending should register online at https://creative.gryphontechnologies.com/darpa/tto/air/pd/. Those seeking to attend should fax their security clearances and visit requests to Amentum at (571) 428-4358. Registration closes on 4 Nov. 2022.

Related: Researchers ask industry for enabling technologies in artificial intelligence (AI) and machine automation

Those attending may meet individually with the Air program manager, Lt. Col. Ryan "Hal" Hefron, on Tuesday 15 Nov. 2022. Email DARPA-SN-2306@ darpa.mil to request an individual session.

One-on-one meetings will be at DARPA at 675 North Randolph St. in Arlington, Va., and will require security clearance. Fax clearance/visit requests to DARPA at (703) 528-3655 or send via encrypted e-mail to VWC@darpa.mil.

Email questions or concerns to Lt. Col. Ryan Hefron at DARPA-SN-23-06@darpa.mil. More information is online at https://sam.gov/opp/1b972abff6de4a2fbf7999af316e52c0/view.

The rest is here:
Military researchers to brief industry on artificial intelligence (AI), sensors, and autonomy program - Military & Aerospace Electronics

AI is not going to replace physicians, but physicians who use AI will replace those who don't.

MIAMI, Oct. 18, 2022 /PRNewswire/ --The new book "How AI Can Democratize Healthcare" by Michael Ferro and Robin Farmanfarmaian dives into the cutting edge of technology moving care from the clinic to where the patient is located, including their home, office, or even traveling. Predictive software, AI voice technology, and digital therapeutics are just some of the innovations detailed in this provocative new title that will shape tomorrow's future today.

Whether people are sick or well, AI-based software programs can monitor conditions in a real-world environment and can be a daily presence in someone's life, delivering personalized interventions precisely when needed. The incremental cost to provide an AI software program to each additional person is negligible so it can scale quickly, and has no geographical boundaries, making worldwide adoption almost effortless.

New ways to improve patient outcomes are in high demand and the major healthcare stakeholders are innovating to bring much needed improvements. AI advancements across Vocal Biomarkers, Remote Patient Monitoring, Digital Therapeutics, Voice Recognition, Decision Support Tools, Virtual Reality, and Predictive Care are converging together to create Ambient Healthcare Computing: the ever-present healthcare assistant that monitors, analyzes, and provides the right interventions to the right person at the right time.

Simply put, AI is revolutionizing the Where, When, What, and How people access healthcare. Legacy systems composed of trained healthcare professionals and physical clinics are a limited and expensive resource. With location removed from the equation, shifting care to the point of the patient increases access and affordability, in effect democratizing healthcare. Further, the negligible incremental cost to provide an AI-based software program per person makes AI infinitely scalable and accessible.

About the Authors:

Healthcare tech entrepreneur and author Michael Ferro is the Founder and CEO Merrick Ventures, a Miami-based PE Firm focused on AI and democratizing healthcare. Michael has built multiple companies that he took public or were acquired for over $1B, including Merge Healthcare and Click Commerce. With the Michael and Jacqueline Ferro Foundation, he has donated millions, including $2M to Northwestern for entrepreneurship and $1M to theMelanoma Research Alliance(MRA). Ferro has won more than 15 awards, including Forbes Tech's 100 Highest Rollers, the Technology Entrepreneur of the Year from Ernst & Young, and was a Nominee for an Emmy Award for Best Documentary.

Professional speaker and entrepreneur Robin Farmanfarmaian has given over 180 talks in 15 countries on technology in healthcare. As an entrepreneur, she has worked with over 20 companies in pharma, device, and AI focused on major diseases including oncology, neuro, diabetes, and more. A misdiagnosis as a teenager led to 43 hospitalizations, six major surgeries, and multiple organs removed. At age 26, Robin fired her healthcare team and took control of her health, including taking herself off high dose opioids. She rebuilt her care team, was diagnosed correctly, and went into remission overnight with the right medication. "How AI Can Democratize Healthcare" is a followup book to Robin's 2015 book, "The Patient as CEO: How Tech Empowers the Healthcare Consumer."

Advance Praise for "How AI Can Democratize Healthcare"

"Artificial Intelligence is going to very rapidly transform medicine, along with ubiquitous cameras and all kinds of sensors. If you want to understand this disruption to our healthcare, Michael and Robin will explain how this will impact you as well as positively impacting billions of lives." Ray Kurzweil, inventor, author, and futurist

"How AI Can Democratize Healthcare'' by Michael Ferro and Robin Farmanfarmaian is easy to read and chock-full of great examples of startups in healthcare AI. We're in the first inning of AI in healthcare, and this book points to a very exciting future. Congratulations on a great book!" John E. Kelly III, PhD, IBM EVP - Retired

"There is overwhelming complexity at the intersection of artificial intelligence, medical technology, human factors and regulatory context. There are almost too many new health care products, services and companies to sort through. Thankfully, How AI Can Democratize Healthcare gives us the old-fashioned sort of intelligence enhancement, clear prose, that makes our collective future both understandable and optimistic. A must read for citizens, investors and policy-makers alike." Bing Gordon, General Partner & Chief Product Officer, Kleiner Perkins

SOURCE Robin Farmanfarmaian

More here:
New Book Heralds a New Era in Healthcare with Artificial Intelligence Already Transforming the Patient Experience - PR Newswire