Mediaboss Marketing

Output of an Artificial Intelligence system from Google Vision, performing Facial Recognition on a ... [+] photograph of a man in San Ramon, California, November 22, 2019. (Photo by Smith Collection/Gado/Getty Images)

The news that Ukraine is using facial recognition software to uncover Russian assailants and identify Ukrainians killed in the ongoing war is noteworthy largely because its one of few documented uses of artificial intelligence in the conflict. A Georgetown University think tank is trying to figure out why while advising U.S. policymakers of the risks of AI.

The CEO of the controversial American facial recognition company Clearview AI told Reuters that Ukraines defense ministry began using its imaging software Saturday after Clearview offered it for free. The reportedly powerful recognition tool relies on artificial intelligence algorithms and a massive quantity of image training data scraped from social media and the internet.

But aside from Russian influence campaigns with their much-discussed deep fakes and misinformation-spreading bots, the lack of known tactical use (at least publicly) of AI by the Russian military has surprised many observers. Andrew Lohn isnt one of them.

Lohn, a senior fellow with Georgetown Universitys Center for Security and Emerging Technology, works on its Cyber-AI Project, which is seeking to draw policymakers attention to the growing body of academic research showing that AI and machine-learning (ML) algorithms can be attacked in a variety of basic, readily exploitable ways.

We have perhaps the most aggressive cyber actor in the world in Russia who has twice turned off the power to Ukraine and used cyber-attacks in Georgia more than a decade ago. Most of us expected the digital domain to play a much larger role. Its been small so far, Lohn says.

We have a whole bunch of hypotheses [for limited AI use] but we dont have answers. Our program is trying to collect all the information we can from this encounter to figure out which are most likely.

They range from the potential effectiveness of Ukrainian cyber and counter-information operations, to an unexpected shortfall in Russian preparedness for digital warfare in Ukraine, to Russias need to preserve or simplify the digital operating environment for its own tactical reasons.

All probably play some role, Lohn believes, but just as crucial may be a dawning recognition of the limits and vulnerability of AI/ML. The willingness to deploy AI tools in combat is a confidence game.

Junk In, Junk Out

Artificial intelligence and machine learning require vast amounts of data, both for training and to interpret for alerts, insights or action. Even when AI/ML have access to an unimpeded base of data, they are only as good as the information and assumptions which underlie them. If for no other reason than natural variability, both can be significantly flawed. Whether AI/ML systems work as advertised is a huge question, Lohn acknowledges.

The tech community refers to unanticipated information as Out of Distribution data. AI/ML may perform at what is deemed to be an acceptable level in a laboratory or in otherwise controlled conditions, Lohn explains. Then when you throw it into the real world, some of what it experiences is different in some way. You dont know how well it will perform in those circumstances.

In circumstances where life, death and military objectives are at stake, having confidence in the performance of artificial intelligence in the face of disrupted, deceptive, often random data is a tough ask.

Lohn recently wrote a paper assessing the performance of AI/ML when such systems scoop in out of distribution data. While their performance doesnt fall off quite as quickly as he anticipated, he says that if they operate in an environment where theres a lot of conflicting data, theyre garbage.

He also points out that the accuracy rate of AI/ML is impressively high but compared to low expectations. For example, image classifiers can work at 94%, 98% or 99.9% accuracy. The numbers are striking until one considers that safety-critical systems like cars/airplanes/healthcare devices/weapons are typically certified out to 5 or 6 decimal points (99.999999%) accuracy.

Lohn says AI/ML systems may still be better than humans at some tasks but the AI/ML community has yet to figure out what accuracy standards to put in place for system components. Testing for AI systems is very challenging, he adds.

For a start, the artificial intelligence development community lacks a test culture similar to what has become so familiar for military aerospace, land, maritime, space or weapons systems; a kind of test-safety regime that holistically assesses the systems-of-systems that make up the above.

The absence of such a back end combined with specific conditions in Ukraine may go some distance to explain the limited application of AI/ML on the battlefield. Alongside it lies the very real vulnerability of AI/ML to the compromised information and active manipulation that adversaries already to seek to feed and to twist it.

Bad Data, Spoofed Data & Classical Hacks

Attacking AI/ML systems isnt hard. It doesnt even require access to their software or databases. Age-old deceptions like camouflage, subtle visual environment changes or randomized data can be enough to throw off artificial intelligence.

As a recent article in the Armed Forces Communications and Electronics Associations (AFCEA) magazine noted, researchers from Chinese e-commerce giant Tencent managed to get a Tesla sedans autopilot (self-driving) feature to switch lanes into oncoming traffic simply by using inconspicuous stickers on the roadway. McAfee Security researchers used similarly discreet stickers on speed limit signs to get a Tesla to speed up to 85 miles per hour in a 35 mile-an-hour zone.

An Israeli soldier is seen during a military exercise in the Israeli Arab village of Abu Gosh on ... [+] October 20, 2013 in Abu Gosh, Israel. (Photo by Lior Mizrahi/Getty Images)

Such deceptions have probably already been examined and used by militaries and other threat actors Lohn says but the AI/ML community is reluctant to openly discuss exploits that can warp its technology. The quirk of digital AI/ML systems is that their ability to sift quickly through vast data sets - from images to electromagnetic signals - is a feature that can be used against them.

Its like coming up with an optical illusion that tricks a human except with a machine you get to try it a million times within a second and then determine whats the best way to effect this optical trick, Lohn says.

The fact that AI/ML systems tend to be optimized to zero in on certain data to bolster their accuracy may also be problematic.

Were finding that [AI/ML] systems may be performing so well because theyre looking for features that are not resilient, Lohn explains. Humans have learned to not pay attention to things that arent reliable. Machines see something in the corner that gives them high accuracy, something humans miss or have chosen not to see. But its easy to trick.

The ability to spoof AI/ML from outside joins with the ability to attack its deployment pipeline. The supply chain databases on which AI/ML rely are often open public databases of images or software information libraries like GitHub.

Anyone can contribute to these big public databases in many instances, Lohn says. So there are avenues [to mislead AI] without even having to infiltrate.

The National Security Agency has recognized the potential of such data poisoning. In January, Neal Ziring, director of NSAs Cybersecurity Directorate, explained during a Billington CyberSecurity webinar that research into detecting data poisoning or other cyber attacks is not mature. Some attacks work by simply seeding specially crafted images into AI/ML training sets, which have been harvested from social media or other platforms.

According to Ziring, a doctored image can be indistinguishable to human eyes from a genuine image. Poisoned images typically contain data that can train the AI/ML to misidentify whole categories of items.

The mathematics of these systems, depending on what type of model youre using, can be very susceptible to shifts in the way recognition or classification is done, based on even a small number of training items, he explained.

Stanford cryptography professor Dan Boneh told AFCEA that one technique for crafting poisoned images is known as the fast gradient sign method (FGSM). The method identifies key data points in training images, leading an attacker to make targeted pixel-level changes called perturbations in an image. The modifications turn the image into an adversarial example, providing data inputs that make the AI/ML misidentify it by fooling the model being used. A single corrupt image in a training set can be enough to poison an algorithm, causing misidentification of thousands of images.

FGSM attacks are white box attacks, where the attacker has access to the source code of the AI/ML. They can be conducted on open-source AI/ML for which there are several publicly accessible repositories.

You typically want to try the AI a bunch of times and tweak your inputs so they yield the maximum wrong answer, Lohn says. Its easier to do if you have the AI itself and can [query] it. Thats a white box attack.

If you dont have that, you can design your own AI that does the same [task] and you can query that a million times. Youll still be pretty effective at [inducing] the wrong answers. Thats a black box attack. Its surprisingly effective.

Black box attacks where the attacker only has access to the AI/ML inputs, training data and outputs make it harder to generate a desired wrong answer. But theyre effective at producing random misinterpretation, creating chaos Lohn explains.

DARPA has taken up the problem of increasingly complex attacks on AI/ML that dont require inside access/knowledge of the systems being threatened. It recently launched a program called Guaranteeing AI Robustness against Deception (GARD), aimed at the development of theoretical foundations for defensible ML and the creation and testing of defensible systems.

More classical exploits wherein attackers seek to penetrate and manipulate the software and networks that AI/ML run on remain a concern. The tech firms and defense contractors crafting artificial intelligence systems for the military have themselves been targets of active hacking and espionage for years. While Lohn says there has been less reporting of algorithm and software manipulation, that would be potentially be doable as well.

It may be harder for an adversary to get in and change things without being noticed if the defender is careful but its still possible.

Since 2018, the Army Research Laboratory (ARL) along with research partners in the Internet of Battlefield Things Collaborative Research Alliance, looked at methods to harden the Armys machine learning algorithms and make them less susceptible to adversarial machine learning techniques. The collaborative developed a tool it calls Attribution-Based Confidence Metric for Deep Neural Networks in 2019 to provide a sort of quality assurance for applied AI/ML.

Despite the work, ARL scientist Brian Jalaian told its public affairs office that, While we had some success, we did not have an approach to detect the strongest state-of-the-art attacks such as [adversarial] patches that add noise to imagery, such that they lead to incorrect predictions.

If the U.S. AI/ML community is facing such problems, the Russians probably are too. Andrew Lohn acknowledges that there are few standards for AI/ML development, testing and performance, certainly nothing like the Cybersecurity Maturity Model Certification (CMMC) that DoD and others adopted nearly a decade ago.

Lohn and CSET are trying to communicate these issues to U.S. policymakers not to dissuade the deployment of AI/ML systems, Lohn stresses, but to make them aware of the limitations and operational risks (including ethical considerations) of employing artificial intelligence.

Thus far he says, policymakers are difficult to paint with a broad brush. Some of those Ive talked with are gung-ho, others are very reticent. I think theyre beginning to become more aware of the risks and concerns.

He also points out that the progress weve made in AI/ML over the last couple of decades may be slowing. In another recent paper he concluded that advances in the formulation of new algorithms have been overshadowed by advances in computational power which has been the driving force in AI/ML development.

Weve figured out how to string together more computers to do a [computational] run. For a variety of reasons, it looks like were basically at the edge of our ability to do that. We may already be experiencing a breakdown in progress.

Policymakers looking at Ukraine and at the world before Russias invasion were already asking about the reliability of AI/ML for defense applications, trying to gauge the level of confidence they should place in it. Lohn says hes basically been telling them the following;

Self driving cars can do some things that are pretty impressive. They also have giant limitations. A battlefield is different. If youre in a permissive environment with an application similar to existing commercial applications that have proven successful, then youre probably going to have good odds. If youre in a non-permissive environment, youre accepting a lot of risk.

The rest is here:
The Vulnerability of AI Systems May Explain Why Russia Isn't Using Them Extensively in Ukraine - Forbes

On a positive note, he says AI has been identified as a key enabler on 79 per cent (134 targets) of the United Nations Sustainable Development Goals (SDGs). However, 35 per cent (59 targets) may experience a negative impact from AI.

Unfortunately, he says unless we start to address the inequities associated with the development of AI right now, were in grave danger of not achieving the UNs SDG goals and, more pertinently, if AI is not properly governed and proper ethics are applied from the beginning, it will have not only a negative physical impact, it will also have a significant social impact globally.

There are significant risks to human dignity and human autonomy, he warns.

If AI is not properly governed and its not underpinned by ethics, it can create socio-economic inequality and impact on human dignity.

A part of the problem at present is most AI is being developed for a commercial outcome, with estimates suggesting its commercial worth to be $15 trillion a year by 2030.

Unfortunately, the path were on poses some significant challenges.

Samarawickrama says AI ethics is underpinned by human ethics and the underlying AI decision-making is driven by data and a hypothesis created by humans.

The danger is much AI is built off the back of the wrong hypothesis because there is an unintentional bias built into the initial algorithm. Every conclusion the AI is making is reached from the hypothesis, which means every decision and the quality of that decision its making is based off a humans ethics and biases.

For Samarawickrama, this huge flaw in AI can only be rectified if diversity, inclusion and socio-economic inequality are taken into account from the very beginning of the AI process.

We can only get to that point if we ensure we have good AI governance and ethics.

The alternative is were basically set up to fail if we do not have that diversity of data.

Much of his work in Australia is with the Australian Red Cross and its parent the International Federation of Red Cross and Red Crescent Societies (IFRC), where he has built a framework linking AI to the seven Red Cross principles in a bid to link AI to the IFRCs global goal of mitigating human suffering.

And while this is enhancing the data literacy across the Red Cross, it also has a potential usage in many organisations, because its about increasing diversity and social justice around AI.

Its a complex problem to solve because there are lot of perspectives as to what mitigating human suffering involves. It goes beyond socio-economic inequality and bias.

For example, the International Committee of the Red Cross is concerned about autonomous weapons and their impact on human suffering.

Samarawickrama says if we are going to achieve the UNSDGs as well as reap the benefits of a $15 trillion a year global economy by 2030, we have to work hard to ensure we get AI right now by focussing on AI governance and ethics.

If we dont, we create a risk of failing to achieve those goals and we need to reduce those by ensuring AI can bring the benefits and value it promises to all of us.

Its why the Red Cross is a good place to start because its all about reducing human suffering, wherever its found and, we need to link that to AI, Samarawickrama says.

Excerpt from:
Award-winner warns of the failures of artificial intelligence - The Australian Financial Review

The earliest form of writing originated nearly 5000 years ago in Mesopotamia (present-day Iraq), representing the Sumerian language. However, these early manuscripts, inscriptions, manuals have suffered the wrath of time. Historians have long worried about the missing texts that could give an insight into the life and culture of ancient civilisation, Artificial Intelligence has now come to their aid.

Named after the Greek island in Homers Odyssey, Ithaca, the first deep neural network will help in not only restoring the missing text of damaged inscriptions, but also identifying their original location, and establishing the date they were written. Designed to assist and expand the historians workflow, this AI has achieved 62 per cent accuracy when restoring damaged texts and improved the accuracy of historians from 25 per cent to 72 per cent.

In a study published in the journal Nature, researchers said that models such as Ithaca can unlock the cooperative potential between artificial intelligence and historians, transformationally impacting the way that we study and write about one of the most important periods in human history.

Inspired by biological neural networks, deep neural networks can discover and harness intricate statistical patterns in vast quantities of data. Ithaca is one such development that merges the fields of technology, supercomputing, and ancient history to reveal unknown secrets hidden in plain sight.

Ithaca was trained to simultaneously perform the tasks of textual restoration, geographical attribution, and chronological attribution. Researchers trained the system on inscriptions written in the ancient Greek language and across the ancient Mediterranean world between the seventh century BC and the fifth century AD.

Credit: Ca' Foscari University of Venice

The architecture of Ithaca was carefully tailored to each of the three epigraphic tasks, meaningfully handling long-term context information and producing interpretable outputs to enhance the potential for human-machine cooperation. We believe machine learning could support historians to expand and deepen our understanding of ancient history, just as microscopes and telescopes have extended the realm of science Yannis Assael, Staff Research Scientist at DeepMind said in a statement.

Researchers said that as centuries went by, many ancient inscriptions were damaged and became partially or completely illegible. In some cases, they were removed from their original location, and they can be difficult to date. For instance, 2500 years ago, Greeks started writing on stone, ceramics, and metal, in order to register all sorts of transactions, laws, calendars, and oracles. Today, these archaeological findings reveal a lot of information on the Mediterranean area. Unfortunately, this tale is incomplete.

DeepMind has partnered with Google Cloud and Google Arts & Culture to launch a free interactive version of Ithaca. (File Pic)

Historians have already used Ithaca to shed light on current disputes in Greek history, including the dating of a series of important Athenian decrees thought to have been written before 446/445 BCE. Ithacas average predicted date for the decrees is 421 BCE, aligning with the new evidence and demonstrating how machine learning might contribute to historical debates.

Although it might seem like a small difference, this date shift has significant implications for our understanding of the political history of Classical Athens. We hope that models like Ithaca can unlock the cooperative potential between AI and the humanities, transformationally impacting the way we study and write about some of the most significant periods in human history, Thea Sommerschield, Marie Curie Fellow at Ca' Foscari University of Venice and fellow at Harvard Universitys CHS said.

Historians are now working on other versions of the AI, which has been trained in different ancient languages to study other ancient writing systems, from Akkadian to Demotic and Hebrew to Mayan.

Go here to see the original:
Meet Ithaca, Artificial Intelligence that will reveal hidden secrets of ancient civilisations - India Today

The JCO is an international, peer-reviewed medical journal published by the American Society of Clinical Oncology (ASCO), with an impact factor (IF) of 44.54. This is the first time that research on AI biomarkers has been published in an international SCI-grade journal of JCO's prestige.

"Immune phenotyping of tumor microenvironment is a logical biomarker for immunotherapy, but objective measurement of such would be extremely challenging," said Professor Tony Mok from the Chinese University of Hong Kong, co-senior author of the journal. "This is the first study that adopted AI technology to define the tumor immune phenotype, and to demonstrate its ability in predicting treatment outcomes of anti-PD-L1 therapy in two large cohorts of patients with advanced non-small cell lung cancer."

Immune checkpoint inhibitors (ICI) are a standard therapy method for advanced NSCLC with programmed death ligand-1 (PD-L1) expression. However, outcomes vary depending on the patient's tumor microenvironment.

Assessing the PD-L1 tumor proportion score (TPS) can bring predictive benefit for patients with high expression (over 50%), who show superior response to ICI therapy over standard chemotherapy. However, ICIs lose their potency in patients with PD-L1 TPS between 1% and 49%, showing outcomes similar to chemotherapy. Therefore, the development of an accuracy-enhanced biomarker to predict ICI response in NSCLC patients with low PD-L1 expression is highly warranted.

While tumor infiltrating lymphocytes (TIL) are promising biomarkers for predicting ICI treatment outcomes apart from PD-L1, clinical application remains challenging as TIL quantification involves a manual evaluation process bound to practical limitations of interobserver bias and intensive labor. Employing AI's superhuman computational capabilities should open new possibilities for the objective quantification of TIL.

To validate immune phenotyping as a complementary biomarker in NSCLC, researchers divided 518 NSCLC patients into three groups based on their tumor microenvironment: inflamed, immune-excluded, and immune-desert. As a result, clinical characteristics based on each immune phenotype group showed statistically significant differences in progression-free survival (PFS) and overall survival (OS).

Furthermore, analysis of NSCLC patients with PD-L1 TPS between 1% and 49% based on their immune phenotype found that the inflamed group showed significantly higher results in objective response rate (ORR) and progression-free survival (PFS), compared to the non-inflamed groups. This shows Lunit SCOPE IO's ability to supplement PD-L1 TPS as a biomarker by accurately predicting immunotherapy response for patients with low PD-L1 TPS.

"Lunit has demonstrated through several abstracts the credibility of Lunit SCOPE IO as a companion diagnostic tool to predict immunotherapy treatment outcomes," said Chan-Young Ock, Chief Medical Officer at Lunit. "This study is a proof-of-concept that compiles all of our past research that elucidates Lunit AI's ability to optimize cancer treatment selection."

Last year, Lunit announced a strategic investment of USD 26 million from Guardant Health, Inc., a leading precision oncology company. Following this major collaboration intended to reshape and innovate the precision oncology landscape, Lunit continues to refine its global position by validating the effectiveness of its AI technology through various studies.

SOURCE Lunit

Read the rest here:
Breakthrough Study Validates Artificial Intelligence as a Novel Biomarker in Predicting Immunotherapy Response - Published in Journal of Clinical...

The European approach to artificial intelligence (AI) will help build a resilient Europe for the Digital Decade where people and businesses can enjoy the benefits of AI. It focuses on 2 areas: excellence in AI and trustworthy AI. The European approach to AI will ensure that any AI improvements are based on rules that safeguard the functioning of markets and the public sector, and peoples safety and fundamental rights.

To help further define its vision for AI, the European Commission developed an AI strategy to go hand in hand with the European approach to AI. The AI strategy proposed measures to streamline research, as well as policy options for AI regulation, which fed into work on the AI package.

The Commission published its AI package in April 2021, proposing new rules and actions to turn Europe into the global hub for trustworthy AI. This package consisted of:

Fostering excellence in AI will strengthen Europes potential to compete globally.

The EU will achieve this by:

The Commission and Member States agreed boost excellence in AI by joiningforces on AI policy and investment. The revised Coordinated Plan on AI outlines a vision to accelerate, act, and align priorities with the current European and global AI landscape and bring AI strategy into action.

Maximising resources and coordinating investments is a critical component of the Commissions AI strategy. Through the Digital Europe and Horizon Europe programmes, the Commission plans to invest 1 billion per year in AI. It will mobilise additional investments from the private sector and the Member States in order to reach an annual investment volume of 20 billion over the course of the digital decade.

The newly adopted Recovery and Resilience Facility makes 134 billion available for digital. This will be a game-changer, allowing Europe to amplify its ambitions and become a global leader in developing cutting-edge, trustworthy AI.

Access to high quality data is an essential factor in building high performance, robust AI systems. Initiatives such as the EU Cybersecurity Strategy, the Digital Services Act and the Digital Markets Act, and the Data Governance Act provide the right infrastructure for building such systems.

Building trustworthy AI will create a safe and innovation-friendly environment for users, developers and deployers.

The Commission has proposed 3 inter-related legal initiatives that will contribute to building trustworthy AI:

The Commission aims to address the risks generated by specific uses of AI through a set of complementary, proportionate and flexible rules. These rules will also provide Europe with a leading role in setting the global gold standard.

This framework gives AI developers, deployers and users the clarity they need by intervening only in those cases that existing national and EU legislations do not cover. The legal framework for AI proposes a clear, easy to understand approach, based on four different levels of risk: unacceptable risk, high risk, limited risk, and minimal risk.

Read this article:
A European approach to artificial intelligence | Shaping ...