Mediaboss Marketing

Discover a better theory

In recent years, the theory of human decision making has skyrocketed. However, these theories are often difficult to distinguish from each other and offer less improvement in explaining decision-making patterns than previous theories.Peterson et al. Leverage machine learning to evaluate classical decision theory, improve predictability, and generate new theories of decision making (see Perspectives by Bhatia and He). This method affects the generation of theory in other areas.

Science, Abe2629, this issue p. 1209abi7668, p. See also. 1150

Predicting and understanding how people make decisions is a long-standing goal in many areas, along with a quantitative model of human decision-making that informs both social science and engineering research. did. Show how large datasets can be used to accelerate progress towards this goal by enhancing machine learning algorithms that are constrained to generate interpretable psychological theories. .. Historical discoveries by conducting the largest experiments on risky choices to date and analyzing the results using gradient-based optimizations of differentiable decision theory implemented via artificial neural networks. A new, more accurate model of human decision-making in the form of summarizing, confirming that there is room for improvement of existing theories, and preserving insights from centuries of research.

Discover the theory of human decision-making using extensive experimentation and machine learning

Source link Discover the theory of human decision-making using extensive experimentation and machine learning

Go here to see the original:
Discover the theory of human decision-making using extensive experimentation and machine learning - Illinoisnewstoday.com

The modern business world is littered with examples where organisations hastily rolled out artificial intelligence (AI) and machine learning (ML)solutions without due consideration of ethical issues, which has led to very costly and painful learning lessons. Internationally, for example, IBM is getting sued afterallegedly misappropriating data from an appwhile Goldman Sachs is under investigation for using anallegedly discriminatory AI algorithm. A closer homegrown example was theRobodebtdebacle, in which the federal governmentdeployed ill-thought-through algorithmic automationtosend out letters torecipientsdemanding repayment ofsocial security payments dating back to 2010. The government settled a class action against it late last year at an eye-watering cost of $1.2 billion after theautomated mailoutssystemtargeted many legitimate social security recipients.

Thattargeting of legitimate recipientswas clearly illegal, says UNSW Business Schools Peter Leonard, a Professor of Practice for the School of Information Systems & Technology Management and the School of Management and Governance at UNSW Business School. Government decision-makersare required by law to take into accountallrelevant considerationsand only relevant considerations, andauthorising automated demands to be made of legitimate recipients was notproper application ofdiscretionsbyan administrative decision-maker.

Prof. Leonard saysRobodebtis an important example of what can go wrong with algorithms in which due care and consideration is not factored in. When automation goeswrong,it usually does soquicklyandat scale. And when things go wrong at scale, you dont need each payout to be much for it to be a very large amount when added together acrossacohort.

Robodebt is an important example of what can go wrong with systems that have both humans and machines in a decision-making chain. Photo: Shutterstock

Technological developments are very often ahead of both government laws and regulations as well as organisational policies around ethics and governance. AI and ML are classic examples ofthisand Prof. Leonard explains there is major translational work to be done in order to bolster companies ethical frameworks.

Theres still a very large gap between government policymakers, regulators, business, and academia. I dont think there are many people today bridging that gap, he observes. It requires translational work, with translation between those different spheres of activities and ways of thinking. Academics, for example, need to think outside their particular discipline,departmentor school. And they have to think about how businesses and other organisations actually make decisions, in order to adapt their view of what needs to be done to suit the dynamic and unpredictable nature of business activity nowadays.Soit isnt easy, but it never was.

Prof. Leonard says organisations are feeling their way to betterbehaviourin this space. Hethinksthat manyorganisationsnow care about adverse societal impacts of their business practices, butdontyet know how to build governance and assurance to mitigate risks associated with data and technology-driven innovation.They dont know how to translate what are often pretty high-level statementsaboutcorporate social responsibility,goodbehaviouror ethics call it what you will into consistently reliable action,to give practical effect to those principles in how they make their business decisions every day. That gap creates real vulnerabilities for many corporations, he says.

Data privacy serves as an example of what should be done in this space. Organisations have become quite good at working out how to evaluate whether a particular form of corporatebehaviouris appropriately protective of the data privacy rights of individuals. This is achieved through privacy impact assessments which are overseen by privacy officers, lawyers and other professionals who are trained to understand whether or not a particular practice in the collection and handling of personal information about individuals may cause harm to those individuals.

Theres an example of how what can be a pretty amorphous concept a breach of privacy is reduced to something concrete and given effect through a process that leads to an outcome with recommendations about what the business should do, Prof. Leonard says.

When things go wrong with data, algorithms and inferences, they usually go wrong at scale. Photo: Shutterstock

Disconnects also exist between key functional stakeholders required to make sound holistic judgements around ethics in AI and ML. There is a gap between the bit that is the data analytics AI, and the bit that is the making of the decision by an organisation. You can have really good technology and AI generating really good outputs that are then used really badly by humans, and as a result, this leads to really poor outcomes, says Prof. Leonard. So, you have to look not only at what the technology in the AI is doing, but how that is integrated into the making of the decision by an organisation.

This problem exists in many fields. Onefieldin which it is particularly prevalent is digital advertising. Chief marketing officers, for example, determine marketing strategies that are dependent upon the use of advertising technology which are in turn managed by a technology team. Separate to this is data privacy which is managed by a different team, and Prof. Leonard says each of these teamsdontspeak the same language as each other in order to arrive at a strategically cohesive decision.

Some organisations are addressing this issue by creating new roles, such as a chief data officer or customer experience officer, who is responsible for bridging functional disconnects in applied ethics. Such individuals will often have a background in or experience with technology, data science and marketing, in addition to a broader understanding of the business than is often the case with the CIO.

Were at a transitional point in time where the traditional view of IT and information systems management doesnt work anymore, because many of the issues arise out of analysis and uses of data, says Prof. Leonard. And those uses involve the making of decisions by people outside the technology team, many of whom dont understand the limitations of the technology in the data.

Why regulatorsneedteeth

Prof. Leonardwas recently appointed to theNSW inaugural AI Government Committee the first of its kind for any federal, state or territory government in Australiatoadvise the NSW Minister for Digital VictorDominelloon how todeliver on key commitments in the states AI strategy.One focusfor the committee ishow to reliably embed ethics in how, when and why NSW government departments and agencies useAIand other automation in their decision-making.

Prof. Leonard said governmentsand other organisationsthat publish aspirational statements and guidance on ethical principles of AIbut fail to go furtherneed to do better.For example, theFederal Governmentsethics principlesforuses ofartificial intelligenceby public and private sector entitieswere publishedover18 months ago, but there is little evidence of adoption across the Australian economy, or that these principles are being embedded into consistently reliable and verifiable business practices, he said.

What good is this? Itis like the 10 commandments.Theyarea great thing. But are people actually going to follow them? And what are we going to do if they dont?Prof. Leonard said it is notworth publishing statements of principles unlessthey are supplemented withprocesses and methodologies for assurance and governance of all automation-assisted decision-making. It is not enough to ensure that the AI component is fair, accountable and transparent: the end-to-end decision-making process must be reviewed.

Technological developments and analytics capabilities usually outpace laws, regulatory policy, audit processes and oversight frameworks. Photo: Shutterstock

While some regulation willalsobe needed to build the right incentives,Prof. Leonard saidorganisations need to first know how to assure good outcomes, before they are legally sanctionedand penalisedfor bad outcomes.The problem for the public sector is more immediate than for the business and not for profit sectors, because poor algorithmic inferences leading to incorrect administrative decisions can directly contravenestate andfederaladministrative law, he said.

In the business and not for profit sectors, thelegalconstraints are more limitedin scope (principally anti-discriminationandscope consumer protection law). Because the legal constraints are limited, Prof. Leonard observed, reporting oftheRobodebtdebacle has not led tosimilarurgency in the business sector asthat inthefederal government sector.

Organisations need to be empowered to thinkmethodically across andthroughpossible harms, whilethere alsoneeds to be adequate transparency in the system and government policy and regulators should not lag too far behind.A combination of these elements will help reduce the reliance on ethics within organisations internally, as they are provided with a strong framework for sound decision-making.And then you come behind with a big stick iftheyrenot using the tools or theyre not using the tools properly. Carrots alone and sticks alone never work; you need the combination of two, said Prof.Leonard.

The Australian Human Rights Commissionsreport on human rights and technologywas recently tabled in Federal Parliament.Human Rights Commissioner EdSantowstatedthat the combination oflearningsfromRobodebtand the Reports findings provide aonce-in-a-generationchallenge and opportunity to develop the proper regulations around emerging technologies tomitigate the risks around them and ensure they benefit all members of the community. Prof Leonard observed that the challenge is as much to how we govern automation aided decision making within organisations the human elementas it is to how we assure that technology and data analytics are fair, accountable and transparent.

Many organisations dont have the capabilities to anticipate when outcomes will be unfair or inappropriate with automation-assisted decision making. Photo: Shutterstock

A good example of the need for this can be seen in the Royal Commission into Misconduct in the Banking, Superannuation and Financial Services Industry. It noted key individuals who assess and make recommendations in relation to prudential risk within banks are relatively powerless compared to those who control profit centres. So, almost by definition, if you regard ethics and policing of economics as a cost within an organisation, and not an integral part of the making of profits by an organisation, you willend up with bad results because you dont value highly enough the management of prudential, ethical or corporate social responsibility risks, says Prof. Leonard. You name me a sector, and Ill give you an example of it.

While he notes that larger organisations will often fumble their way through to a reasonably good decision, another key risk exists among smaller organisations. They dont have processes around checks and balances and havent thought about corporate social responsibility yet becausetheyre not required to, says Prof. Leonard. Small organisations often work on the mantra of moving fast and breaking things and this approach can have a very big impact within a very short period of time,thanks to the potentially rapid growth rate of businesses in a digital economy.

Theyre the really dangerous ones, generally. This means the tools that you have to deliver have to be sufficiently simple and straightforward that they are readily applied, in such a way that an agile move fast and break things' type-business will actually apply them and give effect to thembefore they break things that really can cause harm, he says.

See the original post here:
How to avoid the ethical pitfalls of artificial intelligence and machine learning - UNSW Newsroom

A new papertitled, Cetacean Translation Initiative: a roadmap to deciphering the communication of sperm whales', explains how the scientists are going to try to decode whale vocalisations. The researchers are using machine learning techniques to try and translate the clicking and other noises made by sperm whales, to see if we can understand what the giant creatures are saying.

Whether known non-human communication systems exhibit similarly rich structure either of the same kind as human languages, or completely new remains an open question, reads the concluding sentence in the introduction of the paper, posted on to the preprint server arXiv.org. The paper has been authored by 16 scientific members of Project CETI collaboration.

It was only in the 1950s that we, humans, observed sperm whales made sounds. It took another two decades to understand for humans that they were using those sounds to communicate, according to the new research posted by CETI.

Researchers say that the past decade witnessed a ground-breaking rise of machine learning for human language analysis, and recent research has shown the promise that such tools may also be used for analysing acoustic communication in nonhuman species.

"We posit that the machine learning will be the cornerstone of the future collection, processing, and analysis of multimodal streams of data in animal communication studies," read the abstract of the paper.

And to further understand this, scientists have picked sperm whales, for their highly-developed neuroanatomical features, cognitive abilities, social structures, and discrete click-based encoding, making for an excellent starting point for advanced machine learning tools that can be applied to other animals in the future.

The paper is basically a roadmap towards this goal, they add. Scientists have outlined key elements needed for the collection and processing of massive bioacoustics data of sperm whales, detecting their basic communication units and language-like higher-level structures, and validating these models through interactive playback experiments.

They further say that technological advancements achieved during this effort are expected to help in the application of broader communities investigating non-human communication and animal behavioural research.

Researchers explain that the clicking sound sperm whales make, it appears, serves the dual purpose of echolocation at the depths to which they go and also use it in their social vocalisations. The communication clicks are more tightly packed, according to the CETI paper.

That a project as large as this one would have complexities and challenges is something not very difficult to understand.

David Gruber, a marine biologist, and CETI project leader said that figuring out what they have been able to discover thus far has been challenging, adding, sperm whales have "been so hard for humans to study for so many years." But now, "we actually do have the tools to be able to look at this more in-depth in a way that we haven't been able to before," he said adding, tools included AI, robotics, and drones.

A report in Live Science said that the CETI project has a massive stash of recordings of about 1 lakh sperm whale clicks, painstakingly gathered by marine biologists over many years. However, it said that the machine-learning algorithms might need somewhere close to 4 billion clicks before they start making any conclusions.

And to ensure this, CETI is setting up innumerable automated channels to collect recordings from sperm whales. The tools CETI is using include underwater microphones placed in waters frequented by sperm whales, microphones that can be dropped by eagle-eyed airborne drones as soon as they spot a pod of sperm whales gathering at the surface, and even robotic fish that can follow and listen to whales from a distance, the report said.

If you think collecting these sounds is the only challenge, then wait. According to a 2016 research in the journal Royal Society Open Science, sperm whales are known to have dialects as well. But finding answers to these questions is what CETI is dedicated to.

Go here to read the rest:
Can Humans Ever Understand What Sperm Whales say? This Research Has Roadmap Towards It - Gadgets 360

Pod Save America is actually one undeniably successful example of Democratic progressive messaging, but itlike most avowedly progressive mediais for self-identified politics junkies. Fox News and Sinclair Broadcasting are for anyone with a TV. For much of their audiences, they are simply the news. In order to give voters the positive messages Democrats want them to receive, liberals would need to create a mass media of their own and stop outsourcing the job to the frequently hostile corporate media. The Democrats messaging problem is really a media problem.

Some political science professors summarized a recent research experiment in Politico Magazine earlier this month. Alexander Coppock, Donald P. Green, and Ethan Porter conducted a series of randomized experiments to test whether parties can win over new loyalists with ads that promoted a party rather than a particular candidate. What they found was that, with repeat exposure, people changed their partisan identification ever so slightly after seeing the ads, and that higher doses of party-promoting ads could influence peoples voting decisions and feelings about Donald Trump. Partisan identity is usually understood as a root cause of political behavior, the political scientists wrote. By moving it, we also appear to have moved real-world political decisions.

In the world of American political communications, ads promoting a party are a novelty. The researchers concluded that both parties could benefit from producing the kinds of ads we tested, and its true that neither party currently does this with conventional TV advertising. But while these political scientists framed their experiments as part of a novel ad strategy, what they were really doing was directly exposing people to particular political messages that had been designed to influence their political affinitiesand even their identities. There is already language to describe what that kind of messaging is. These political scientists independently invented party propaganda, exposed Americans to it, and discovered that it can be effective, especially with constant exposure. Conservatives dont need to learn to do this: Its how their movement sustains itself.

Amusingly, the top-shelf political ad professionals the political scientists hired to make the ads were flummoxed by the request, because no one had ever before asked them to create messaging designed to promote the Democratic Party or to convince people to associate themselves with it. Despite how familiar American liberals are with the power of propaganda when yielded by the right, it has seemingly never occurred to the most powerful of them to do any propagandizing on behalf of their own causes and party!

Read this article:
Why Liberals Ought to Invest in Propaganda - The New Republic

Philippe J. Fournier: The latest federal election projection shows the Liberals falling short of a majority, with an outcome eerily similar to the results of the 2019 election

If rumours swirling around Parliament Hill in Ottawa are true, then Canadians will be called to vote in the 44th Canadian federal election late this summer or early next fall, just days before this Parliament blows its second candle in October. Crunching the numbers over the weekend, the only question that kept popping in my head is: Why?

Both the Bloc Qubcois and the NDP hold the balance of power in a Liberal-lead Parliament, hence neither party should feel any hurry to rush Canadians to the polls since both parties find themselves at their height of relevance. Time can feel like an eternity as a third or fourth party in a majority parliament.

Therefore, should an election be called, it will most likely entirely be a Liberal initiative. Cynics will say, not necessarily wrongly, that a fall election would be power-grab attempt by the Liberals: The numbers are currently favourable for the Liberal Party, and it would obviouslymuch rather spend the next four years in a majority position, not having to worry about pesky opposition MPs asking questions or making occasional threats to vote against the government in a confidence vote.

Recent precedent would also stand on the side of an election in September or October. John Horgan, Blaine Higgs and Andrew Fureyall premiers leading their respective legislature with minority status one year agochose to bet on their provinces relatively good performance at handling the pandemic, and all three were re-elected with majorities. With a significant fraction of Canadians projected to be fully vaccinated by late August, the national discussion could gradually move away from the pandemic and dive into the handling of the economic recovery, so the Liberals could understandably be tempted to cash in some karma chips before the inevitable What have you done for me lately? feeling sets in. Politicians know that voters memory tends to be short.

However, are current numbers that good for the Trudeau Liberals? Perhaps the partys internal polling is showing different trends than those of polls released for public consumption, because this weeks 338Canada federal update measures the most likely outcome as being eerily similar to the results of the 2019 federal election.

Federal polls published in the past month have shown the Liberals leading the Conservatives by margins between one and 11 points, with a current average of five points. With such numbers, the Liberals would almost assuredly win the most seats if an election were held this week, but the party would most likely end up short of the 170-seat threshold for a majority at the House of Commons. Here are this weeks averages per party:

The Liberals win an average of 163 seats, seven short of majority status, but only six seats above their 2019 results. While the Liberals remain dominant in Atlantic Canada and continue to lead in seat-rich Ontario, their only potential seat gains as currently projected would be found in Quebec, where the LPC averages 38 per cent and 41 seats. Nonetheless, let us remember that the Liberals won 35 Quebec seats in 2019 (and 40 in 2015). Given that the Bloc Qubcois support remains in relatively good shape (just under the 30 per cent mark), it is rather unlikely that the Liberals can find many more seats to gain in the province. As for Atlantic Canada, adding Fredericton from the Greens doesnt hurt (with Jenica Atwin crossing the floor from the Greens to the Liberals this week), but its unlikely to have any effect beyond this electoral districts border.

As for the Conservatives, they appear stuck in a high-floor and low-ceiling scenario that would almost assuredly guarantee them Official opposition status. In the past month, the Conservatives have polled between 27 and 32 per cent nationally, and have shown no significant gain in Central Canada where the party needs it most. If it cannot grow its own support beyond the current numbers, the only scenario in which the party wins a plurality of seats would be if the NDP outperforms its polls and expectations.

To wit: The NDPs current polling average in Ontario stands at 20 per cent, three points higher than its 2019 result in the province. Should the NDPs vote in the next federal actually match its polling results, a net gain of six to 12 seats would be entirely plausibleand most of those seats would come at the expense of the Liberals. In the past week alone, boththe Angus Reid InstituteandLgermeasured NDP support above the 20 per cent mark nationally, and even had the NDP getting the support of one in four Ontario voters (25 per cent from Angus Reid and 24 per cent from Lger). With such numbers, a complete 25-seat sweep of Toronto would be almost impossible for the Liberals (unlike in both 2015 and 2019 federal elections). Without a harvest of Ontario seats similar to those of 2015 and 2019 for the LPC, a majority would simply be mathematically out of reach for Justin Trudeau.

Naturally, the projection confidence intervals do stretch into majority territory for the Liberals. In these scenarios, the Liberals would have to outperform their current standings and hope the NDP fails to effectively get out its vote, especially in Ontario. In the waning days of the 2019 campaign, the polling average showed the NDP at 18 per cent nationally, yet it ended up with 16 per cent. This modest, but measurable two per cent gap probably cost the NDP a dozen seats from coast to coast, most of them in Ontario.We cannot discount the possibility that the NDP would fail to match its improving poll results, especially if the NDP relies on its support from younger voters (who tend to vote in lesser numbers).

Hence, with NDP support still hovering just under 20 per cent and the Bloc Qubcois still riding the CAQs coattails in Quebec, where will the Liberals find enough seats to secure a majority? Perhaps an additional seat in Manitoba? Perhaps one seat in each of Edmonton and Calgary? Maybe retake Vancouver-Granville from independent MP Jody Wilson-Raybould? Nunavut? All of these perhaps and maybes add up to a lot of uncertainty, and make for a rather poor risk to reward ratio.

Follow 338Canada on Twitter

* * *

Details of this projection are available on the 338Canada page. To find your home district, use the list of all 338 electoral districts here, or use the regional links below:

Read more:
A summer election would be a risky bet for the Liberals: 338Canada - Maclean's