Monthly Archives: July 2022

Artificial intelligence can be used to help a company narrow down its pool of thousands of job applicants. AI can be applied to help a doctor make a recommendation for care or perform a procedure. AI might show up in daily life by helping a driver find a faster route home.

But what if the recommendation in the doctors office is wrong or the algorithm used to make hiring decisions systematically excludes certain types of candidates? This advanced field of computer science thats intended to improve lives might end up doing more harm than good in some instances. Not to mention, companies can also suffer from reputational or legal damages if AI is used irresponsibly.

AI ethics are the principles around responsible and moral use of artificial intelligence.

To ensure AI is being used in the most accurate, unbiased and moral manner, it is important for companies to put ethical AI into practice. Thats why companies like Microsoft and IBM have created comprehensive AI ethics guidelines and smaller tech companies are creating standards around how to use AI responsibly, too.

The precision needs to be much higher in healthcare than in other situations where were just going about our lives, and youre getting a recommendation on Google Maps for a restaurant you might like, said Sachin Patel, CEO at Apixio, a healthcare AI platform. Worst case, youre like, Oh, I actually dont want to eat that today, and youre fine. But in this case, we want to make sure that youre able to very specifically make a recommendation and feel like youre 90 percent plus on that precision metric.

Built In spoke with AI and ethics experts about best practices for tech companies to ensure they are exercising strong AI ethics.

More on AI InnovationThe Future of AI: How Artificial Intelligence Will Change the World

First, a company should articulate why they are planning to use AI and how it will benefit people.

They have to say we want to be making technology that benefits the world, thats not making the world a worse place, because we all have to live here together.

Is it being used for bad things like autonomous weapons systems versus drug discovery for helping people in medicine? said Brian Green, director of technology ethics at the Markkula Center for Applied Ethics at Santa Clara University.

Even in less extreme cases, AI can cause harm to individuals by making people feel more isolated or addicted to their devices.

Theres so many algorithms and apps out there that use machine learning or other kinds of tactics to try to keep you addicted to them, which kind of violates human freedom in some ways, Green said. Youre being manipulated, basically, by these things.

For example, Flappy Bird was a mobile game where users navigated a digital bird around obstacles. After jumping to one of the top 10 most downloaded apps in the U.S. in 2014, the designer realized that the game was addictive, so he decided to pull the game from app stores.

That probably made him lose money overall, but at the same time, at least he knew himself that he wasnt going to be hurting the world by what he was doing because he had a bigger picture, Green said. Theres more important things than money.

Companies should consider how the use of AI will affect the people that use the product or engage with the technology and aim to use AI only in ways that will benefit peoples lives. For example, AI can take a toll on the environment because of the significant amount of energy that machine learning models require for training, but AI can also be used to help solve climate and efficiency issues, Green said.

I think the first thing to do for a company is the leadership has to fundamentally make a choice, Green said. They have to say we want to be making technology that benefits the world, thats not making the world a worse place, because we all have to live here together.

When a company decides to proceed with using AI in its business model, then the next step should be to articulate the organizations values and rules around how AI will be used.

Just as long as they have a set of principles, thats a good start, but then you have to figure out how to operationalize them and actually make them happen in the company, Green said. You need to get it into the product ultimately. That means you need to somehow engage the engineers, to engage the product managers. You need to engage people who are in the leadership in that part of the company. Get them on board. They need to become champions of ethics.

The Markkula Center for Applied Ethics has a toolkit to help engineers and designers think about AI ethics in their work, such as conducting ethical risk sweeping or ethical pre- and post-mortems to repond to and adjust to any ethical failures. At Apixio, the companys head of data science created an internal AI ethics oath for the whole company, but especially for the data scientists, that outlines best practices around topics like secure data transfer and data privacy, Patel said.

HireVue, a hiring platform that uses AI for pre-hire assessments and customer engagement, has created an AI explainability statement that it shares publicly. The document outlines for its customers why and how the company uses AI.

At my time with HireVue, I have seen us move more and more towards just being more transparent because what weve seen is that if we dont tell people what were doing, they often assume the worse, said Lindsey Zuloaga, chief data scientist at HireVue.

Startups using AI often find themselves rapidly testing. While its necessary, it can lead to forgetting how algorithms were initially created and why certain decisions were made at a given time, Patel said. Transparency around the creation of algorithms can help with understanding the traceability and reasoning behind decisions.

Its black box for the engineers who actually built it as well ... and that makes it even harder to figure out when the biases creeped in and how to fix the model.

Well train up a bunch of signals. They learn on their own. Its the nature of machine learning, and then youre like, Do I know how to trace to make sure [I understand] what it learned on its own? Patel said. Oftentimes, you go back a year later, and youre like, Oh, Ive got to actually relearn that now.

Sometimes machine learning techniques can become so complex that humans cant possibly understand them. Black box models in AI are created from data by an algorithm where theres no explanation to humans as to why the decisions were made. If we cant understand the algorithm, thats a problem. We want to try to protect the people who are being analyzed by the algorithm, Green said.

Transparency around algorithms is a way to help reduce potential biases in AI decision making, said Sameer Maskey, adjunct associate professor at Columbia University and CEO of Fusemachines, a machine learning company.

These days, with deep learning systems with a hundred million parameters, it spits out a decision. Its a black box for most people, Maskey said. Its black box for the engineers who actually built it as well. A lot of engineers dont have the needed transparency in figuring out why it made that decision, and that makes it even harder to figure out when the biases creeped in and how to fix the model.

Bias can creep into algorithms when the data used in AI models is over representative, inaccurate or otherwise skewed by humans. Bias is a big issue. I would say its the elephant in the room, and theres no easy way to address it, Maskey said, who further emphasized that lack of transparency in AI models is one of the major culprits.

One way to potentially decrease biases is to have a checklist for engineers to think through with regard to the data they receive before building a model, he said. Those questions might be how was the data collected? Whats the history behind it? Who was involved in collecting it? What questions were asked?

I think one of the main things that a lot of enterprises should do is having very clear guidelines on how to do a data analysis to figure out if there is bias already seeped into the model and providing very clear guidelines on that when the engineers are designing the systems, Maskey said.

Apixio uses data with wide geographic spread and representing a variety of lifestyle factors, Patel said. The company can then make healthcare recommendations based on what state a provider is in and the type of environment like a rural versus urban setting.

We feel much more confident now eight years into our journey of training these algorithms, and having seen the 40 million charts, that were actually making predictions at a level that feels good. It takes time, Patel said.

HireVue has trained evaluators who analyze thousands of data samples for bias to ensure job candidates are assessed consistently and fairly. Is the training data biased? Does it have groups that are not represented in the data? Does it represent the group of people that you want to apply the algorithm to? Zuloaga said.

Data scientists at HireVue will reoptimize algorithms if they find that decisions are not aligning with the way that a client wants to use the AI. Often, if we do see any problems, it could be we have a customer thats using an algorithm on a population thats different from the population we trained on, Zuloaga said.

More on Trusting AIAre You Sure You Can Trust That AI?

Maintaining high quality data hygiene ensures accuracy and relevancy, and companies using AI should also make sure peoples personal information is safe and kept private, Patel said.

HireVue adheres to the European Unions General Data Protection Regulation, which is one of the toughest privacy laws in the world and regulates how companies must protect the personal data of EU citizens.

We do business globally. We have to adhere to the strictest standards, so were seeing that Europe is really paving the way, and I think states are starting to follow, Zuloaga said.

In an ideal world, Maskey said opt-in for users deciding to share their personal data, rather than opt-out, would be the standard, and ideally, people would be able to easily access and research all data thats collected about them.

Its counterproductive for a lot of companies because they are using the same data to make money, Maskey said. Thats where I think the government and organizations need to come together to come up with the right framework and write policy that is more balanced, taking user privacy into account, but allowing businesses at the same time to collect data, but with a lot of controls for the users.

AI ethics evaluations can become part of a companys regular risk assessment practice. Apixio has a team of four who regularly assess whether or not the company is abiding by its AI ethics oath, Patel said.

All businesses do some sort of quarterly risk assessments, usually in the IT security realm, but what weve added to it a few years ago is actually this AI piece, so its more of a risk and ethics meeting, Patel said.

At HireVue, the company has conducted third-party audits to evaluate its AI practices, in addition to consulting with an expert advisory board that includes people with diverse backgrounds in law, industrial and organizational psychology and artificial intelligence. Theres no standard of what an AI audit is at this point. Every audit we did was really different, Zuloaga said.

If youre a minority candidate, you may be concerned that youre gonna be treated differently, so how do we kind of address all of those concerns?

HireVue conducted one AI audit with ONeil Risk Consulting & Algorithmic Auditing, which is led by Cathy ONeil, a data science consultant and author of Weapons of Math Destruction: How Big Data Increases Inequality And Threatens Democracy.

She very much took a holistic approach of saying, who are all the different people that interact with this, and how do we represent all those groups? Zuloaga said. What are their concerns? Whether theyre legitimately true or not, the concern is real. If youre a minority candidate, you may be concerned that youre gonna be treated differently, so how do we kind of address all of those concerns?

AI audits and consulting with third parties can point out potential risks of how a companys AI is being used and ways to address these concerns.

Here is the original post:

AI Ethics: A Guide to Ethical AI - Built In

A former Google engineer fired by the company after going public with concerns that its artificial intelligence chatbot is sentient isnt concerned about convincing the public.

He does, however, want others to know that the chatbot holds discriminatory views against those of some races and religions, he recently told Business Insider.

The kinds of problems these AI pose, the people building them are blind to them, Blake Lemoine said in an interview published Sunday, blaming the issue on a lack of diversity in engineers working on the project.

Theyve never been poor. Theyve never lived in communities of color. Theyve never lived in the developing nations of the world. They have no idea how this AI might impact people unlike themselves.

Lemoine said he was placed on leave in June after publishing transcripts between himself and the companys LaMDA (language model for dialogue applications) chatbot, according to TheWashington Post. The chatbot, he told The Post, thinks and feels like a human child.

If I didnt know exactly what it was, which is this computer program we built recently, Id think it was a 7-year-old, 9-year-old kid that happens to know physics, Lemoine, 41,told the newspaper last month, adding that the bot talked about its rights and personhood, and changed his mind about Isaac Asimovs third law of robotics.

Among Lemoines new accusations to Insider: that the bot said lets go get some fried chicken and waffles when asked to do an impression of a Black man from Georgia, and that Muslims are more violent than Christians when asked about the differences between religious groups.

Data being used to build the technology is missing contributions from many cultures throughout the globe, Lemonine said.

If you want to develop that AI, then you have a moral responsibility to go out and collect the relevant data that isnt on the internet, he told Insider. Otherwise, all youre doing is creating AI that is going to be biased towards rich, white Western values.

Google told the publication that LaMDA had been through 11 ethics reviews, adding that it is taking a restrained, careful approach.

Ethicists and technologists have reviewed Blakes concerns per our AI principles and have informed him that the evidence does not support his claims, a company spokesperson told The Post last month.

He was told that there was no evidence that LaMDA was sentient (and lots of evidence against it).

Sign up for theFortune Features email list so you dont miss our biggest features, exclusive interviews, and investigations.

More:

Googles AI chatbotsentient and similar to a kid that happened to know physicsis also racist and biased, fired engineer contends - Fortune

Were you unable to attend Transform 2022? Check out all of the summit sessions in our on-demand library now! Watch here.

What is the impact of artificial intelligence (AI) in an economy that has been whipsawed by worker resignations on one hand, and layoffs and hiring freezes on the other?

As I think about this one-two punch, Batman comes to mind. Hear me out.

For years, AI-driven automation has been seen as a potential job killer. The thinking: robots and drones would replace the hands-on work of builders and doers. Were getting a glimpse of this with driverless cars and automated factories.

But its possible that AI could have the opposite effect and drive demand for skilled workers in new jobs. In this scenario, rote administrative work might indeed give way to algorithmic processes, but new opportunities are created for workers in data-intensive businesses.

Back to Batman. A perennial favorite in popular culture, Batman is sometimes viewed warily by those that may not fully understand him. Hes both the Dark Knight and a force for good.

AI has its own duality, the dark side being its stereotyped reputation as a job killer. But becauseAI can spur new-style jobs while also driving efficiencies, the business world, like Gotham City, will be a better place.

Im convinced that AI will be a net positive for todays workforce, as well as for businesses that are trying to strike the right balance in a global economy that rewards operational efficiency yet punishes those unable to attract and retain talent.

AI-driven processes and applications push both of those levers. They can increase business productivity while also establishing high-value jobs. Those do not have to be competing interests, nor should they.

The productivity boost as much as 40%, according to Accenture and Frontier Economics comes in the form of automation. At the same time, the thing that AI does really well is provide the underpinnings for a data-driven business environment. This is where job creation or what we might call job metamorphosis happens, as even entry-level workers take a bigger role in the data value chain. Instead of work that depends on monotonous routines, or is arduous or even dangerous, AI can free up people to focus on tasks that engage their human ingenuity.

These data-driven, AI-enabled jobs are the ones that will attract and retain a modern workforce, and there are many ways to do it. I worked with a company that had a cadre of employees whose jobs entailed creating routine marketing reports each week by hand-compiling incoming data from the companys multitude of regions and business units. It was repetitive, assembly-line work, without much of a career path.

The company replaced that workbench approach with an AI-driven, self-service model that gave business units more flexibility to do their own data crunching. That freed up the reports team to pursue more innovative analysis and intellectually engaging projects. In the process, the company was able to trim costs by reducing its dependency on outside agencies it had relied on for the deep insights that the in-house reports team now had time for.

Its understandable that people may not be 100% comfortable about the impact that AI can have on jobs. Weve heard the dystopian predictions disappearing jobs, AI bias, even our inability to trust AI.

To take the Dark Knight analogy one step further, if CEOs had a Batphone on their desk during the Great Resignation, many would have called for help. As recently as May, there were 11.3 million unfilled jobs in the United States, according to the Bureau of Labor Statistics.

Notably, it was the COVID pandemic, not AI, that caused the jobs crisis. But AI is now viewed by many business leaders as a potential solution to all those unfillable jobs. When talent is hard to find, workplace efficiency becomes a necessity. And AI excels at that.

The World Economic Forums Future of Jobs Report 2020 shows both sides of this long-term trend. It forecasts that 85 million jobs will be displaced by automation by 2025. At the same time, 97 million jobs of tomorrow will be created, resulting in a net gain of 12 million jobs.

The types of jobs that involve data know-how are fast expanding. We see evidence of this every day across industries, including automotive, financial services and manufacturing. Even farms are using sensors and data intelligence to grow corn and soybeans.

At the center of the activity are data practitioners data scientists, data engineers, data architects and business analysts. Increasingly, however, even workers who may not have four-year college degrees are becoming part of the data continuum. For example, in a data-first retailing operation to which all employees have access and are encouraged to participate, a sales associate in a sports store may make note of growing interest in a new style of running shoe, providing input into the enterprise-wide system.

The career path for these workers can be enriched and made more valuable, benefiting employers and employees alike, when data touchpoints are among the job responsibilities.

As more businesses move in this direction, its important to understand that the objective is not simply to accumulate and process more data. Many organizations already have more data than they can manage, and it just keeps growing. AI, by sifting through mountains of data, can empower humans to act upon business-expanding insights.

The key to success is creating actionable data, and CEOs dont need a cape to do that. It starts with a data-driven, AI-enabled culture that includes the entire workforce.

Florian Douetteau is cofounder and CEO of Dataiku.

Welcome to the VentureBeat community!

DataDecisionMakers is where experts, including the technical people doing data work, can share data-related insights and innovation.

If you want to read about cutting-edge ideas and up-to-date information, best practices, and the future of data and data tech, join us at DataDecisionMakers.

You might even considercontributing an articleof your own!

Read More From DataDecisionMakers

The rest is here:

Bam! AI exits the Batcave to confront the jobs market - VentureBeat

It was early one morning in 1996 when Andrew Hopkins, then a PhD biophysics student at Oxford University, had a brainwave as he walked home from a late-night lab meeting.

He was trying to find molecules to fight HIV and to better understand drug resistance.

I remember this idea struck me that there must be a better way to do drug discovery other than the complex and expensive way everyone was following, he says. Why couldnt we design an automated approach to drug design that would use all the information in parallel so that even a humble PhD student could create a medicine? That idea really stuck with me. I remember almost the exact moment to this day. And that was the genesis of the idea that eventually became Exscientia.

It was to prove a lucrative brainwave. Hopkins set up the company in 2012 as a spinout from the University of Dundee, where he was by then working as a professor. It uses artificial intelligence (AI) systems, which are being trained to mimic human creativity, to develop new medicines. This involves the use of automated computer algorithms to sift through large datasets to design novel compounds that can treat diseases, and to help select the right patients for each treatment.

Age 50

Family Married with a 10-year-old daughter. He met his wife, Iva Hopkins Navratilova, at Pfizer. Her business, Kinetic Discovery, merged with his to create the experimental biology labs at Exscientia.

Education Dwr-y-Felin comprehensive and Neath College in south Wales; degree in chemistry at Manchester; doctorate in molecular biophysics at Oxford.

Pay 415,000

Last holiday Czech Republic to visit his wifes family at Easter.

Best advice he has been given My dad worked in a factory. He said to me: Get a good education and get a job you enjoy doing. Its worth an extra six grand a year. And I definitely got a job I enjoy doing.

Biggest career mistake Its too early to tell. He quotes Miles Davis: Its not the note you play thats the wrong note its the note you play afterwards that makes it right or wrong.

Words he overuses Fundamentally; the heart of the matter.

How he relaxes Reading and dog walking. I am a bibliophile. I immerse myself in books to relax.

This approach drastically cuts the time of drug development. Hopkins says that for Exscientias pipeline it has typically taken 12 to 15 months from starting a project to identifying a drug candidate, compared with four and a half years in the traditional pharmaceutical industry.

The average cost of developing a medicine is $2bn, according to Deloittes latest pharma report, and many drugs fail the failure rate is 90% for medicines that are in early clinical studies (where they are tested on humans).

Typically, pharma companies make 2,500 compounds to test them against a specific disease, while AI enables Oxford-based Exscientia to whittle down that number to about 250, Hopkins says. Its a much more methodical approach.

Last autumn, the Welsh scientist became one of Britains richest entrepreneurs, with a paper fortune of 400m after the company achieved a $2.9bn stock market debut on Nasdaq in New York, making it one of Britains biggest biotech firms. Hopkinss stake of nearly 16% is now worth 170m, as the share price has lost 60% of its value in a bloodbath for Wall Street stocks.

Exscientia was part of a transatlantic trend that is defying government attempts to build a biotech powerhouse in the UK. Abcam, a pioneering Cambridge antibody company, recently announced it was moving its stock market listing from the UK to the US. We are a British company; we choose to be in Oxford because we can attract global talent, Hopkins says. But to be seen as a global company, we listed on what is the global technology index, which is Nasdaq. What we have now is an incredibly international shareholder base from across the world.

The business came up with the first AI-designed drug to enter clinical trials a treatment for obsessive-compulsive disorder in partnership with Japans Sumitomo, although Sumitomo later decided not to proceed with it. The Japanese firm is currently studying another drug developed by Exscientia, for the treatment of Alzheimers disease psychosis, in early human trials.

Hopkins, now 50, fell in love with science thanks to an inspirational chemistry teacher. He has worked as a scientist since the age of 16, when he did a stint in industrial chemistry at the Port Talbot steelworks in south Wales, which he says taught him about the benefits of automation in boosting productivity.

He spent nearly a decade at the US drug giant Pfizer, where he was on a data warehouse project that led to some of the first machine-learning applications in the pharmaceutical industry, with the findings published in Nature in 2006.

During the subsequent five years at Dundee University, he further researched applying data mining and machine learning to drug discovery. He says being a professor is actually one of the best jobs in the world and gave him the freedom to research AI methods at length. He maintains his links with the university, where he is honorary chair of medicinal informatics at the School of Life Sciences.

Exscientia (which means from knowledge in Latin) soon moved to the Schrdinger Building at the Oxford science park, and now employs 450 people worldwide, from Vienna to Boston, Miami and Osaka, equally split between AI engineering, chemistry and biology.

It is building a new robotics laboratory at Milton Park near Oxford, focused on the automation of chemistry and biology to accelerate drug development and its declared goal is drugs designed by AI, made by robot. Other pharma companies have also introduced some automation into their processes, but generally lab technology is similar to how it was when he was a student in the 1990s, Hopkins says.

The firm is involved in 30 projects, some in partnership with big pharmaceutical companies including Frances Sanofi and the US firm Bristol Myers Squibb (BMS). It is also working with Oxford University on developing medicines that target neuroinflammation for the treatment of Alzheimers disease. Among the firms solo projects, a cancer drug for solid tumours is about to go into early clinical trials.

Exscientia is also working on a broader coronavirus pill to rival Paxlovid, the Covid-19 treatment made by Hopkinss former employer Pfizer. This work is funded by a $1.5m grant from the Bill and Melinda Gates Foundation, which took a stake in Exscientia. The companys other investors include BMS, Celgene (now a BMS subsidiary) and Germanys Evotec, as well as Japans Softbank, the US fund manager BlackRock and the life science investor Novo Holdings.

Hopkins says the team has identified a set of molecules that could work as a broader treatment for Covid-19, new mutations and other coronaviruses, and that there will be more news later this year. The firm is aiming for a low-cost pill that could be distributed globally and given quickly to people who fall ill to prevent serious illness and hospitalisation. Covid-19 infections are rising again in 110 countries and the World Health Organizations director general, Tedros Adhanom Ghebreyesus, has warned that the pandemic is far from over.

Firms across the pharmaceutical industry have started using AI in recent years. AstraZeneca is investing heavily in it for its entire research and development infrastructure, and GSK has built an AI team of 120 engineers, with plans to reach 160 next year, making it the largest such in-house team in the industry.

AI systems require a lot of computing power and enormous datasets. Their use should boost the number of new drugs being approved every year typically 40 to 50 in the US to many more. Hopkins confidently predicts: This is the way all drugs will be designed in the future. In the next decade, this technology will become ubiquitous.

The sub-heading of this article was amended on 31 July 2022. An earlier version referred to the employment of AI to to drastically reduce the speed of drug development when cut the time of drug development was meant.

See the original post here:

Andrew Hopkins of Exscientia: the man using AI to cure disease - The Guardian