The link between artificial intelligence (AI) and bias is alarming.

As AI evolves to become more human-like, its becoming clear that human bias is impacting technology in negative, potentially dangerous ways.

Here, we explore how AI and bias are linked and whats being done to reduce the impact of bias in AI applications:

See more: The Ethics of Artificial Intelligence (AI)

Using AI in decision-making processes has become commonplace, mostly because predictive analytics algorithms can perform the work of humans at a much faster and often more accurate rate. Decisions are being made by AI on small matters, like restaurant preferences, and critical issues, like determining which patient should receive an organ donation.

While the stakes may differ, whether human bias is playing a role in AI decisions is sure to impact outcomes. Bad product recommendations impact retailer profit, and medical decisions can directly impact individual patient lives.

Vincent C. Mller takes a look at AI and bias in his research paper, Ethics of Artificial Intelligence and Robotics, included in the Summer 2021 edition of The Stanford Encyclopedia of Philosophy. Fairness in policing is a primary concern, Mller says, noting that human bias exists in the data sets used by police to decide, for example, where to focus patrols or which prisoners are likely to re-offend.

This kind of predictive policing, Mller says, relies heavily on data influenced by cognitive biases, especially confirmation bias, even when the bias is implicit and unknown to human programmers.

Christina Pazzanese refers to the work of political philosopher Michael Sandel, a professor of government, in her article, Great promise but potential for peril, in The Harvard Gazette.

Part of the appeal of algorithmic decision-making is that it seems to offer an objective way of overcoming human subjectivity, bias, and prejudice, Sandel says. But we are discovering that many of the algorithms that decide who should get parole, for example, or who should be presented with employment opportunities or housing replicate and embed the biases that already exist in our society.

See more: Artificial Intelligence: Current and Future Trends

To figure out how to remove or at least reduce bias in AI decision-making platforms, we have to consider why it exists in the first place.

Take the AI chatbot training story in 2016. The chatbot was set up by Microsoft to hold conversations on Twitter, interacting with users through tweets and direct messaging. In other words, the general public had a large part in determining the chatbots personality. Within a few hours of its release, the chatbot was replying to users with offensive and racist messages, having been trained on anonymous public data, which was immediately co-opted by a group of people.

The chatbot was heavily influenced in a conscious way, but its often not so clear-cut. In their joint article, What Do We Do About the Biases in AI in the Harvard Business Review, James Manyika, Jake Silberg, and Brittany Presten say that implicit human biases those which people dont realize they hold can significantly impact AI.

Bias can creep into algorithms in several ways, the article says. It can include biased human decisions or reflect historical or social inequities, even if sensitive variables such as gender, race, or sexual orientation are removed. As an example, the researchers point to Amazon, which stopped using a hiring algorithm after finding it favored applications based on words like executed or captured, which were more commonly included on mens resumes.

Flawed data sampling is another concern, the trio writes, when groups are overrepresented or underrepresented in the training data that teaches AI algorithms to make decisions. For example, facial analysis technologies analyzed by MIT researchers Joy Buolamwini and Timnit Gebru had higher error rates for minorities, especially minority women, potentially due to underrepresented training data.

In the McKinsey Global Institute article, Tackling bias in artificial intelligence (and in humans), Jake Silberg and James Manyika lay out six guidelines AI creators can follow to reduce bias in AI:

The researchers acknowledge that these guidelines wont eliminate bias altogether, but when applied consistently, they have the potential to significantly improve on the situation.

See more: Top Performing Artificial Intelligence Companies

Read the original:
Bias in AI: Algorithm Bias in AI Systems to Know About 2021 - Datamation

A major new report on the state of artificial intelligence (AI) has just been released. Think of it as the AI equivalent of an Intergovernmental Panel on Climate Change report, in that it identifies where AI is at today, and the promise and perils in view.

From language generation and molecular medicine to disinformation and algorithmic bias, AI has begun to permeate every aspect of our lives.

The report argues that we are at an inflection point where researchers and governments must think and act carefully to contain the risks AI presents and make the most of its benefits.

The report comes out of the AI100 project, which aims to study and anticipate the effects of AI rippling out through our lives over the course of the next 100 years.

AI100 produces a new report every five years: the first was published in 2016, and this is the second. As two points define a line, this second report lets us see the direction AI is taking us in.

One of us (Liz Sonenberg) is a member of the standing committee overseeing the AI100 project, and the other (Toby Walsh) was on the study panel that wrote this particular report. Members of the panel came from across the world, with backgrounds in computer science, engineering, law, political science, policy, sociology and economics.

The report highlights the remarkable progress made in AI over the past five years. AI is leaving the laboratory and has entered our lives, having a real-world impact on people, institutions, and culture. Read the news on any given day and youre likely to find multiple stories about some new advance in AI or some new use of AI.

For example, in natural language processing (NLP), computers can now analyse and even generate realistic human language. To demonstrate, we asked Open AIs GPT-3 system, one of the largest neural networks ever built, to summarise the AI100 report for you. It did a pretty good job, even if the summary confronts our sense of self by being written in the first person:

In the coming decade, I expect that AI will play an increasingly prominent role in the lives of people everywhere. AI-infused services will become more common, and AI will become increasingly embedded in the daily lives of people across the world.

I believe that this will bring with it great economic and societal benefits, but that it will also require us to address the many challenges to ensure that the benefits are broadly shared and that people are not marginalised by these new technologies.

A key insight of AI research is that it is easier to build things than to understand why they work. However, defining what success looks like for an AI application is not straightforward.

For example, the AI systems that are used in healthcare to analyse symptoms, recommend diagnoses, or choose treatments are often far better than anything that could be built by a human, but their success is hard to quantify.

Read more: GPT-3: new AI can write like a human but don't mistake that for thinking neuroscientist

As a second example of the recent and remarkable progress in AI, consider the latest breakthrough from Googles DeepMind. AlphaFold is an AI program that provides a huge step forward in our ability to predict how proteins fold.

This will likely lead to major advances in life sciences and medicine, accelerating efforts to understand the building blocks of life and enabling quicker and more sophisticated drug discovery. Most of the planet now knows to their cost how the unique shape of the spike proteins in the SARS-CoV-2 virus are key to its ability to invade our cells, and also to the vaccines developed to combat its deadly progress.

The AI100 report argues that worries about super-intelligent machines and wide-scale job loss from automation are still premature, requiring AI that is far more capable than available today. The main concern the report raises is not malevolent machines of superior intelligence to humans, but incompetent machines of inferior intelligence.

Once again, its easy to find in the news real-life stories of risks and threats to our democratic discourse and mental health posed by AI-powered tools. For instance, Facebook uses machine learning to sort its news feed and give each of its 2 billion users an unique but often inflammatory view of the world.

Its clear were at an inflection point: we need to think seriously and urgently about the downsides and risks the increasing application of AI is revealing. The ever-improving capabilities of AI are a double-edged sword. Harms may be intentional, like deepfake videos, or unintended, like algorithms that reinforce racial and other biases.

AI research has traditionally been undertaken by computer and cognitive scientists. But the challenges being raised by AI today are not just technical. All areas of human inquiry, and especially the social sciences, need to be included in a broad conversation about the future of the field. Minimising negative impacts on society and enhancing the positives requires consideration from across academia and with societal input.

Governments also have a crucial role to play in shaping the development and application of AI. Indeed, governments around the world have begun to consider and address the opportunities and challenges posed by AI. But they remain behind the curve.

A greater investment of time and resources is needed to meet the challenges posed by the rapidly evolving technologies of AI and associated fields. In addition to regulation, governments also need to educate. In an AI-enabled world, our citizens, from the youngest to the oldest, need to be literate in these new digital technologies.

At the end of the day, the success of AI research will be measured by how it has empowered all people, helping tackle the many wicked problems facing the planet, from the climate emergency to increasing inequality within and between countries.

AI will have failed if it harms or devalues the very people we are trying to help.

The rest is here:
Artificial intelligence is now part of our everyday lives and its growing power is a double-edged sword - The Conversation AU

Its time for your Farm of the Future Report. Im Tim Hammerich.

Often when we hear that artificial intelligence is going to change agriculture, the explanation of how thats going to look is a bit fuzzy. We are starting to see examples though in the way of automation. Precision AI is a Canadian startup thats developing drone spraying technology that is currently being trialed on farms. CEO Daniel McCann says its an example of how artificial intelligence can open up new ways of farming.

McCann The previous paradigm that's basically time immemorial is again, because human beings don't have the ability to make per plan level decisions on a giant field, you're going to spray everything, right. But with automation and AI well now everybody like, it's not just us, there's other people building robots, capable of making per plant level decisions. And when you can make a per plant level decision, it's a complete game changer, right? Because now as you say, we don't need to worry about whether or not this particular chemical's impact on the crop is going to be problematic because you won't have to spray the crop. Right? So even things that aren't currently possible, like using some organics like agricultural vinegar to control weeds, well you can't use that on a crop today, but if you could precisely spray it and target just the plants, now that becomes actually a viable method of weed control and resistance control. So there's all sorts of new types of completely outside the box approaches to this problem that I think are going to become the defacto standard as this technology becomes more and more prominent. And it's inevitable, it's coming, the efficiency improvements are just too great.

Tune in tomorrow to hear more about how artificial intelligence is finding real applications on the farm.

Excerpt from:
Artificial Intelligence Opens New Farming Possibilities - AG INFORMATION NETWORK OF THE WEST - AGInfo Ag Information Network Of The West

Now, there's a lot of methodology that goes into this report. There's also judgment into the weightings of the methodology. Anyone that's ever dealt with an index understands how that works out. At Eurasia Group, we've had political risk indices for decades. And it's very clear that although it's quantitative, it's qualitative too, right?

But in this case, the numbers that were expected by the Chinese government, they were ranked number 78 in the 2017 report. They thought that they were doing better, and it turned out that the ranking was going to go down. It was going to be number 85. The Chinese government was quite surprised about this. The World Bank management was quite surprised about this. Chinese government came back and said, what the hell's going on? And according to Kristalina and Jim Kim, and again this was under Jim Kim at the time, they said, okay, go back, take a serious look and make sure that you didn't make any mistakes.

So far, so good. The Chinese ranking eventually came back to number 78, the same as it was the previous year. In other words, after the complaints, the ranking went up seven points.

Now, was there a methodology problem? Was there undue pressure being placed on the analysts? That's the result of an investigation that was ordered by the World Bank under David Malpass, the new leader of that institution. And Jim Kim is gone, but Kristalina Georgieva has gone on to bigger and better things, now running the IMF.

Now, a couple of interesting things about this report. First of all, when the law firm WilmerHale originally went to the IMF, to Georgieva, and this was, I guess, back in July, and wanted to interview her about the report, they explicitly said in the letter they sent her that she was not a subject of investigation, she was just being called in as a witness. And what I find interesting about all of that is Kristalina wasn't worried about her own role at all. She didn't bring in any lawyers. She didn't take any legal advice. She just went and spoke to them immediately, and later became subject of the investigation. And when the report came out, it's blaming her for involvement.

I think the fact that she chose to simply chat with them, number one implies she's not enormously politically savvy in a way that I think that Christine Lagarde, the former head of the IMF, would've been much more careful and cautious. But it also does show motive that she really didn't believe that there was anything she was personally involved in that would be a problem, and so why wouldn't she just answer their questions? In other words, she had no reason to think that this was going to be a problem for her own job or her own tenure.

Second point, this is a report that I would argue normally wouldn't have an awful lot of international impact, except China is massively politicized. Anything having to do with China these days, influence over the World Health Organization, that's nominally why President Trump then decided to leave the WHO in the middle of a pandemic. I mean, any potential sniff that the United States and an appointee in an institution, a multilateral institution where the US has the most votes, the IMF, was helping the Chinese, they're going to run in the other direction.

So it is problematic because it's China. And it's interesting in this regard, that the economic policy makers in the Biden administration like Georgieva, they think that her policies and her tenure so far have been strong, and they're generally supportive of her. But the political types in the White House think that they should run away, because they do not want any ability of opponents to be able to say you guys supported someone who's in the pocket of the Chinese, how dare you.

And again, in these days where there's zero trust between the United States and China, and it's only politically beneficial to be seen as more of a hawk, whether you're a Democrat or a Republican, you understand why they're doing that. Having said all of that, the Europeans have been supportive of Georgieva, continue to be so after a board meeting where they brought in both WilmerHale, as well as the Managing Director. And certainly if there were a smoking gun, if there were email evidence or other witnesses that were directly involved, and knew that Kristalina had directly and unduly pressured them, that would have come out, and the Europeans wouldn't be supporting her at this point. It's not like they're in her pocket.

So the Americans are backing off of her. The Japanese are with the Americans. It's a new Japanese government. They don't particularly have a dog in the fight, and they tend to line up with the Americans when things are important to the US. That being the case, I would say it's damaging to her tenure, but she sticks it out. I don't think she's going anywhere.

And so that's where I think we are in a nutshell, presently. It's going to be harder for the United States to be as aligned with her going forward. Very interesting thing, under Trump, despite the fact that Lagarde was seen as a multilateralist, and she's a European, she's French, the Trump administration never had a hard time with the IMF. In fact, the relations between Christine and Trump, and particularly former Secretary of the Treasury, Mnuchin, were actually very strong. And the reason for that was because there was a very warm and personal relationship between Ivanka, Trump's daughter, and Christine Lagarde.

And Ivanka went to the White House and basically said don't do anything to this organization, they're important, they're useful. He didn't really care so he left it alone. So it wasn't politicized under Trump. It's getting a little politicized under Biden because of the China issue, who would've expected that? But that's where we are, that's what I think, and that's your Quick Take today. Everyone be good, I'll talk to y'all real soon.

Read this article:
Podcast: The future of artificial intelligence with tech CEO Kai-Fu Lee - GZERO Media

InA Citizens Guide to Artificial Intelligence,John Zerilli, John Danaher, James Maclaurin, Colin Gavaghan, Alistair Knott, Joy Liddicoat and Merel Noorman offer an overview of the moral, political, legal and economic implications of artificial intelligence (AI). Exemplary in the clarity of its explanations, the book provides an excellent foundation for considering the issues raised by the integration of AI into our societies, writes Karl Reimer.

A Citizens Guide to Artificial Intelligence. John Zerilli, John Danaher, James Maclaurin, Colin Gavaghan, Alistair Knott, Joy Liddicoat and Merel Noorman. MIT Press. 2021.

Find this book (affiliate link):

A Citizens Guide to Artificial Intelligence is a text that ought to be read widely. The books subject matter is highly relevant and it provokes many probing questions that deserve further consideration on the part of the reader and broader society.

The book itself covers a multitude of topics, ranging from What is Artificial Intelligence?, where the science behind Artificial Intelligence (AI) is described, to Deep Learning, machine learning, neural networks and other material. Later in the text, Algorithms in Government and Oversight and Regulation consider the integration of AI into daily life. Given space constraints, I will focus on two particular chapters in closer detail: Transparency and Responsibility and Liability.

Authors John Zerilli et al begin the Transparency chapter with an anecdote suggesting one shouldnt trust technologies unless one has a way to investigate them (22). From this, they draw the relevant question: what exactly does it mean for a system to be transparent? The remainder of the chapter is dedicated to expanding upon the various facets of transparency including responsibility, accountability, accessibility and inspectability.

Zerilli et al explain the legal concepts behind the right that an individual has to an explanation. They then contrast this against algorithmic systems that fail to provide reasonable explanations for their actions. Further, such algorithmic systems cannot be appealed in their decisions as is the case in traditional legal cases (28). This then provokes the question of what explanations have been demanded of AI systems before the authors emphasise the relevant topic of standard-setting for AI. A given example of such a standard is the European Unions General Data Protection Regulation (GDPR) (30).

Image Credit:Photo byPossessed PhotographyonUnsplash

The Responsibility and Liability chapter is similar in its approach, which is evidence of the overall readability of the text. In this chapter, the leading questions are: do we want machines to be held responsible for decisions? and can machines be responsible? Drawing upon the work of jurisprudence scholar H.L.A. Hart, a hypothetical scenario of a drunk sea captain is given to highlight the complexity of responsibility (62). In the scenario, a fictional sea captain is responsible for the safety of his passengers and crew. However, he becomes drunk every evening of his voyage and at one point the ship is lost amidst a storm with no survivors. Because of his drunken state at the time of the accident, the sea captain is considered negligent and criminally responsible for the loss of life in legal proceedings following the accident. The important observation is that responsibility is a complex notion, which can refer to causal contribution as well as the obligations and duties that come with the professional role of a sea captain (62). Just as responsibility is complex for the sea captain, so too is it complex for AI machines.

Zerilli et al note that responsibility comes in the form of moral responsibility generally attached to individuals and legal responsibility attached to individuals as well as corporate entities such as Google or Facebook. Between AI technologies and human beings, there seems to be a responsibility gap whereby it is difficult to hold human individuals responsible for actions (71). For example, one can consider how a multitude of programmers combine their efforts to create a singular automated driving system one can argue it would be unreasonable to place responsibility onto an individual programmer. Toward the conclusion of the chapter, Zerilli et al also probe the possibility of morally and legally responsible AI, which indeed deserves consideration given the complexity of AI issues (76).

From a discussion of these two chapters, one can get a sense of the style Zerilli et al have used across the book. It is a well grounded and objective work. The text goes to the point of the actual issues that academics are working on at this moment. For this, the text is insightful and prescient: it neither presents the topic of AI as science fiction nor is it dismissive of the capabilities of AI. The text is also accessible to a wide audience. The authors come from legal and philosophical backgrounds yet are able to describe the complexities of AI systems with subtle nuance. The first chapter where the variety of AI systems is explained exemplifies this.

The criticism I have of the text is that it is uneven in the quality of its content as well as its scope and breadth. Consider the brief introduction, Algorithms in Government. Zerilli et al rightly observe a trade-off between legitimacy and efficiency (130), whereby policy agencies become removed from the citizenry who elected them when they are given increased discretionary unelected power. An example of this is the controversy surrounding UK GCSEs and A Levels based on a systematic calibration conducted by education regulator Ofqual. However, in this instance, as in others, there is little nuance to the argument made, which generally identifies a problem relating to AI, offers case studies showing this and then concludes that we need to think carefully about AI in society.

The Algorithms in Government chapter could be extended to discuss further issues, such as whether an AI system can itself wield any legitimate authority. Further, the chapter on Control could be summarised as an introduction to what scholars call the control problem when and whether humans ought to delegate effective control to AI systems. It is a reflection of the thin spread of the book that this one problem has an entire chapter for its discussion as opposed to being integrated into questions of responsibility and legitimacy.

A Citizens Guide to Artificial Intelligence is nonetheless a text that deserves to be read widely. It offers a sufficiently broad overview of the expansive literature on AI. Its a book that one could recommend to any individual without feeling guilty about sharing an overly complex topic. Zerilli et al are exemplary in the clarity of their explanations of AI and its influence on society.

The takeaway is this AI as we conceive of it is already integrated in our society through the algorithms and automated decisions carried out by policymakers and corporate entities. As such, a deep consideration of these issues is necessary. The work put forward by Zerilli et al is an excellent foundation to this end.

Please read our comments policy before commenting.

Note: This article gives the views of the author, and not the position of USAPP American Politics and Policy, nor of the London School of Economics.

Shortened URL for this post:https://bit.ly/3DHOssH

Karl Reimer Philosophy ExchangeKarl Reimer graduated from the London School of Economics with an Msc. in Philosophy and Public Policy. He now supports the Philosophy Exchange project (https://philosophyexchange.org/about/), aimed at sharing philosophical ideas and fostering community within the discipline via regular podcasts, conferences and a vibrant international community.

Read the original:
Book Review: A Citizen's Guide to Artificial Intelligence by John Zerilli, John Danaher, James Maclaurin, Colin Gavaghan, Alistair Knott, Joy...