Cloud Hosting

Artificial intelligence (AI) needs to be more accountable but ethical considerations are not keeping pace with the technologys rate of deployment, says a panel of experts.

This is partly due to the black box nature of AI, whereby its almost impossible to determine how or why an AI makes the decisions it does, as well as the complexities of creating an unbiased AI.

However, according to panelists at the Bristol Technology Showcase, transparency is not enough, with greater accountability being the key to solving many of the ethical issues surrounding AI.

Meaningful transparency doesnt simply follow from doing things like open sourcing the code, thats not sufficient, says Eamonn ONeill, professor of computer science at the University of Bath and director of the UKRI Centre for Doctoral Training in Accountable, Responsible and Transparent AI.

Code and deep learning networks can be opaque however hard you try to open them to inspection. How does seeing a million lines of code help you understand what your smartphones mid-ware is doing? Probably not a lot.

ONeill says that AI needs to be accompanied by a chain of accountability that holds the systems human operator responsible for the decisions of the algorithm.

We dont go to a company and say I cant tell if youve cooked the books because I cant access the neurons of your accountants nobody cares about accountants neurons, and we shouldnt care about the internal workings of AI neural networks either, he said.

Instead, ONeill says we should be focusing on outcomes.

John Buyers, chair of the AI and Ethics panel and a partner at law firm Osborne Clark, points to the example of Mount Sanai Hospital using an AI system called Deep Patient, which was made to trawl through thousands of electronic health records.

Over the course of doing that, Deep Patient became very adept at diagnosing, among other things, adult schizophrenia, which human doctors simply couldnt do, he says. They dont know how the system got to that, but it was of demonstrable public benefit.

Zara Nanu, CEO of human resources technology company Gapsquare, says: When we talk about bias, its bias in terms of the existing data we have that machines are looking at, but also the bias in algorithms we then apply to the data.

She gives the example of Amazon, which gathered a team of data scientists to develop an algorithm that would help it identify top engineers from around the world, who could then be recruited by the company.

All was going well except the machines had learnt to exclude women from the candidate pool, so it was down-scoring people who had woman on their CV, and it was actually scoring higher people who had words like lead or manage, she says.

Amazon came under scrutiny and tried to look how they could make it fairer, but they had to scrap the programme because they couldnt hand-on-heart say the algorithm wouldnt end up discriminating against another group.

Therefore, while accountability does not remove potential bias in the first place, it did make Amazon, as the entity operating the AI system, responsible for the negative effects or consequences of that bias.

However, Chris Ford, a Smith and Williamson partner responsible for a $270m AI investment fund, says theres a critical deficit in the way many corporate entities are approaching the deployment of the technology.

MIT Sloan and Boston Consultancy Group produced an interesting paper earlier this year surveying 3,000 companies globally, most of them outside North America, he says.

What was eye catching was that of those who responded, about half of them said they can see no strategic risk in the deployment of AI platforms within their business, and I find that quite extraordinary.

Ford says this is partly due to a fear of missing out on the latest technological trends, but also because there is not enough emphasis on ethics in education related to AI.

He notes the example of Stuart Russells book,Artificial intelligence: A modern approach, which has been through numerous iterations and is one of the most popular course texts in the world.

That textbook in its most recent form is up to 1,100 pages, he says. Its extraordinarily comprehensive, but the section that deals with ethics is covered in the first 36 pages.

So theres an issue on emphasis here, both in respect to the academic training of data scientists but also what theyre expected to engage with in the commercial world when they leave education.

In terms of bias, the panelists also note that what is socially normal or acceptable is biased in itself.

The question then becomes whose societal norms are we talking about? We are already seeing significant differences and perspectives in the adoption of AI in different parts of the world, says Ford.

Buyers summarised that a lack of bias is not the introduction of objectivity, but the application of subjectivity in accordance with societal norms, so its incredibly difficult.

The overall argument is that AI, like humans, will always be biased to a point of view, meaning transparency will only go so far in solving the ethical issues around the deployment of AI.

Using AI in contrast to humans can facilitate transparency we can fully document the software engineering process, the data, the training, the system performance these measures can be used to support systematic inspection, and therefore transparency and regulation, but accountability and responsibility must stay with the humans, says ONeill.

The Bristol Technology Showcase was held in November 2019, and focused on the impact of emerging technologies on both businesses and wider society.

Link:

Accountability is the key to ethical artificial intelligence, experts say - ComputerWeekly.com

Artificial intelligence is now commonly used in journalism for anything from combing through large datasets to writing stories.

To help you prepare for the future, the Journalism AI team at Polis, London School of Economics and Political Science (LSE), put together a training module seven things to consider before adopting AI in your news organisation.

"Keep in mind that this is not a manual for implementation," writes professor Charlie Beckett who leadsJournalism AI.

"The recommendations will help you reflect on your newsroom AI-readiness but they wont tell you how to do design a strategy. We link to more resources that might help you with that and we hope to produce more training resources ourselves in the near future."

For more insights into the Journalism AI report, you can watch this three-minute video, as well as Charlie Becketts presentation of the report at its launch event.

If you like our news and feature articles, you can sign up to receive our free daily (Mon-Fri) email newsletter (mobile friendly).

Read the rest here:

Tip: Seven recommendations for introducing artificial intelligence to your newsroom - Journalism.co.uk

Editors Note:All opinion section content reflects the views of the individual author only and does not represent a stance taken by The Collegian or its editorial board. Letters to the Editor reflect the view of a member of the campus community and are submitted to the publication for approval.

To the Editor,

I am writing this essay to bring awareness and recognition to a fast-approaching topic in the field of military technology weaponized artificial intelligence.

Weaponized AI is any military technology that operates off a computer system that makes its own decisions. Simply put, anything that automatically decides a course of action against an enemy without human control would fall under this definition.

Weaponized AI is a perfect example of a sci-fi idea that has found its way into the real world and is not yet completely understood. This said, weaponized AI places global security at risk and must be recognized by institutions like Colorado State University before it becomes widely deployed on the battlefield.

Nations are constantly racing to employ the next best weapon as it is developed. AI is no exception. Currently, AI is responsible for the one of the largest technology competitions since that of nuclear weapons during the Cold War. At the top of this competition is China and the United States.

With little to no international restrictions on the deployment of AI weaponry, a modern arms race will continue to develop, creating tensions between world powers as fear of the opposing team reaching the perfect AI weapon arises.

The other inherent danger is the gap that is being created between advanced world powers and countries who are incapable of developing such technology. The tendency for global conflict to occur between these nations increases, as powers that wield weaponized AI have a distinct edge over countries that do not employ AI. This allows room for misuse of this power given the lack of international regulations on using this tech.

What we have is a blurring of moral boundaries as we come closer to allowing this technology to determine who is a true threat.

Going further, my studies have shown that this technology poses considerable risk to international human rights laws. In its current state, weaponized AI is found to be unreliable in doing what it is intended to do. As an example, Project Maven, a current AI used by the United States, only identifies military threats using complex algorithms.

While this seems harmless, the direction in which the world is taking this technology is not. What would happen if this technologys unreliability costs innocent lives due to a targeting error that AIs, like Project Maven, are prone to making? Likewise, who would take responsibility for the actions of a machine?

What we have is a blurring of moral boundaries as we come closer to allowing this technology to determine who is a true threat. These kinds of errors cannot be tolerated by the rules of modern warfare.

A final obstacle surrounding AI is the United Nations inability to come to a consensus on its use. Researcher Eugenio Garcia with the United Nations stated, Advanced military powers remain circumspect (guarded) about introducing severe restrictions on the use of these technologies.

Although people easily recognize the dangers that AI poses to national security, countries are not willing to restrict the development. Furthermore, with minimal current legislation on the unreliability of the technology, weaponized AI will move further than what we can control.

While I make these claims, one must recognize that the technology does offer the benefit of removing soldiers from the battlefield. However, nations around the world are not monitoring this rising issue.

Colorado State University, being a tier one research facility that has investment in military technology, will be the institution that does step up to the plate and recognize catastrophe before it happens. These threats to global security may not be present now, but if we do not advocate for international legislation, these dangers will become reality.

Sincerely,

Thomas Marshall

Third-year mechanical engineering student at CSU

Working under Azer Yalin as an undergraduate research assistant exploring Air Force technology

The Collegians opinion desk can be reached atletters@collegian.com. To submit a letter to the editor, pleasefollow the guidelines at collegian.com.

Original post:

LTTE: It's important to know of weaponized artificial intelligence - Rocky Mountain Collegian

PARIS--(BUSINESS WIRE)--Iktos, a company specialized in Artificial Intelligence for novel drug design and Almirall, S.A. (ALM), a leading skin-health focused global pharmaceutical company, today announced a collaboration agreement in Artificial Intelligence (AI), where Iktos generative modelling technology will be used to design novel optimized compounds, to speed up the identification of promising drug candidates for undisclosed Almirall drug discovery program(s).

Iktos AI technology, based on deep generative models, helps bring speed and efficiency to the drug discovery process, by automatically designing virtual novel molecules that have all desirable characteristics of a novel drug candidate. This tackles one of the key challenges in drug design: rapid and iterative identification of molecules which simultaneously validate multiple bioactive attributes and drug-like criteria for clinical testing.

This partnership is an example of how we intend to explore the enormous possibilities offered by technology to find new molecules and to speed up clinical development, said Dr. Bhushan Hardas, Executive Vice President R&D, Chief Scientific Officer of Almirall. The health sector lags behind others in the digital world. Almirall wants to be at the forefront of innovation to develop holistic and transversal approaches. Artificial Intelligence will provide Almirall a unique opportunity to combine our proficiency with the preciseness and celerity to truly make a difference in patients' lives.

We are thrilled to initiate a new research collaboration with Almirall commented Yann Gaston-Math, President and CEO of Iktos. This new collaboration is further testimony to the leadership position that Iktos has developed in the field of AI for de novo drug design, in little more than two years of existence. We are eager to demonstrate to our collaborators the power of Iktos technology to accelerate their research, and to get the opportunity to further improve by confronting our approach to a new use case, consistently with our strategy to prove our value in real-life projects.

Iktos has recently announced several collaborations with biopharmaceutical companies where Iktos AI technology is used to accelerate the design of promising compounds, and has published, at the EFMC 2018 meeting, an experimental validation of the technology in a real-life drug discovery project. Iktos generative modelling SaaS software, Makya, is now available on the market, and Iktos intends to release its retrosynthesis SaaS platform Spaya as a beta version, before the end of 2019.

About Iktos

Incorporated in October 2016, Iktos is a French start-up company specialized in the development of artificial intelligence solutions applied to chemical research, more specifically medicinal chemistry and new drug design. Iktos is developing a proprietary and innovative solution based on deep learning generative models, which enables, using existing data, to design molecules that are optimized in silico to meet all the success criteria of a small molecule discovery project. The use of Iktos technology enables major productivity gains in upstream pharmaceutical R&D. Iktos offers its technology both as professional services and as a SaaS software platform, Makya.

About Almirall

Almirall is a leading skin-health focused global pharmaceutical company that partners with healthcare professionals, applying Science to provide medical solutions to patients and future generations. Our efforts are focused on fighting against skin health diseases and helping people feel and look their best. We support healthcare professionals by continuous improvement, bringing our innovative solutions where they are needed.

The company, founded almost 75 years ago with headquarters in Barcelona, is listed on the Spanish Stock Exchange (ticker: ALM). Almirall has been key in value creation to society according to its commitment with to major shareholders and through its decision to help others, to understand their challenges and to use Science to provide solutions for real life. Total revenues in 2018 were 811 million euros. More than 1,800 employees are devoted to Science.

For more information, please visit almirall.com

Read the original:

Iktos and Almirall Announce Research Collaboration in Artificial Intelligence for New Drug Design - Business Wire