UMass Amherst and Stanford team designs new framework for fairer, safer algorithms
Seventy years ago, science fiction writer Isaac Asimov imagined a world where robots would serve humans in countless ways, and he equipped them with built-in safeguards now known as Asimov’s Three Laws of Robotics, to prevent them, among other goals, from ever harming a person.
Guaranteeing safe and fair machine behavior is still an issue today, says machine learning researcher and lead author Philip Thomas at the University of Massachusetts Amherst. “When someone applies a machine learning algorithm, it’s hard to control its behavior,” he points out. This risks undesirable outcomes from algorithms that direct everything from self-driving vehicles to insulin pumps to criminal sentencing, say he and co-authors.
Writing in Science, Thomas and his colleagues Yuriy Brun, Andrew Barto and graduate student Stephen Giguere at UMass Amherst, Bruno Castro da Silva at the Federal University of Rio Grande del Sol, Brazil, and Emma Brunskill at Stanford University this week introduce a new framework for designing machine learning algorithms that make it easier for users of the algorithm to specify safety and fairness constraints.
“We call algorithms created with our new framework ‘Seldonian’ after Asimov’s character Hari Seldon,” Thomas explains. “If I use a Seldonian algorithm for diabetes treatment, I can specify that undesirable behavior means dangerously low blood sugar, or hypoglycemia. I can say to the machine, ‘while you’re trying to improve the controller in the insulin pump, don’t make changes that would increase the frequency of hypoglycemia.’ Most algorithms don’t give you a way to put this type of constraint on behavior; it wasn’t included in early designs.”
“But making it easier to ensure fairness and avoid harm is becoming increasingly important as machine learning algorithms impact our lives more and more,” he says.
However, “a recent paper listed 21 different definitions of fairness in machine learning. It’s important that we allow the user to select the definition that is appropriate for their intended application,” he adds. “The interface that comes with a Seldonian algorithm allows the user to do just this: to define what ‘undesirable behavior’ means for their application.”
In Asimov’s Foundation series, Seldon is in the same universe as his Robot series. Thomas explains, “Everything has fallen apart, the galactic empire is collapsing, partly because the Three Laws of Robotics require certainty. With that level of safety required, robots are paralyzed with indecision because they cannot act with certainty and guarantee that no human will be harmed by their actions.”
Seldon proposes fixing this by turning to reasoning probabilistically about safety. “That’s a good fit to what we’re doing, Thomas says. The new approach he and colleagues provide allows for probabilistic constraints and requires the algorithm to specify ways the user can tell it what to constrain. He says, “The framework is a tool for the machine learning researcher. It guides them toward creating algorithms that are easier for users to apply responsibly to real-world problems.”
To test the new framework, they applied it to predict grade point averages in a data set of 43,000 students in Brazil by creating a Seldonian algorithm with constraints. It successfully avoided several types of undesirable gender bias. In another test, they show how an algorithm could improve the controller in an insulin pump while guaranteeing that it would not increase the frequency of hypoglycemia.
Thomas says, “We believe there’s massive room for improvement in this area. Even with our algorithms made of simple components, we obtained impressive results. We hope that machine learning researchers will go on to develop new and more sophisticated algorithms using our framework, which can be used responsibly for applications where machine learning used to be considered too risky. It’s a call to other researchers to conduct research in this space.”
Learn more: New Machine Learning Algorithms Offer Safety and Fairness Guarantees
The Latest Google Headlines on:
Artificial intelligence algorithms
[google_news title=”” keyword=”Artificial intelligence algorithms” num_posts=”10″ blurb_length=”0″ show_thumb=”left”]
The Latest Bing News on:
Artificial intelligence algorithms
- Nvidia Owns a 3.4% Stake in This Innovative Artificial Intelligence (AI) Stock Cathie Wood Loveson April 27, 2024 at 5:58 am
Nvidia's second-largest holding is a 3.4% stake in Recursion Pharmaceuticals ( RXRX 3.57%). The company is using AI to mine data and discover new potential drugs. And if it successfully develops a ...
- 3 Under-the-Radar Artificial Intelligence (AI) Stocks to Buy Nowon April 27, 2024 at 12:15 am
Cybersecurity is a remarkably competitive but crowded market. Eventually, the cream rises to the top, and SentinelOne is among the best security products. It's a regular high-performer on third-party ...
- U.S. Government Needs to ‘Get It Right’ on Artificial Intelligenceon April 26, 2024 at 6:59 pm
We can't afford to get this wrong—again,” Shalanda Young, the director of the Office of Management and Budget, tells TIME.
- Colorado lawmakers consider landmark artificial intelligence regulationson April 26, 2024 at 8:35 am
Colorado could be one of the first states to broadly regulate generative artificial intelligence, as lawmakers consider the balance between setting guardrails on a potentially-harmful technology and ...
- Art Meets Algorithm – Your Guide to AI-Driven Cartoon Magicon April 25, 2024 at 9:22 pm
Imagine stepping into a world where your everyday photos transform into whimsical cartoon versions of themselves, not by the stroke of an artist’s brush, but by the swift algorithms of artificial ...
- The Power of Algorithms: 10 Ways Generative AI is Revolutionizing the Way We Shop Onlineon April 25, 2024 at 12:15 pm
In the dynamic world of online shopping, algorithms are reshaping the way we browse, choose, and purchase products. Generative AI is at the forefront of this revolution, harnessing cutting-edge ...
- Decentralized AI is key to more unbiased AI algorithms — Masa co-founderon April 25, 2024 at 7:00 am
Google’s Gemini AI has faced recent outrage after generating historically inaccurate imagery and biased results.
- Spotting AI Washing: How Companies Overhype Artificial Intelligenceon April 24, 2024 at 10:56 pm
I'll be looking at the phenomenon of AI washing – what it is, who is doing it, why it is dangerous, and perhaps most importantly, how to spot it.
- CT Senate approves bill to regulate artificial intelligence and criminalize ‘deepfake’ pornographyon April 24, 2024 at 3:10 pm
Trying to regulate the fast-moving world of artificial intelligence, the state Senate was still debating Wednesday night on a sweeping bill that creates a framework designed to help consumers and ...
- Do You (Really) Need A Chief Artificial Intelligence Officer?on April 23, 2024 at 1:56 pm
How should organizations manage the tsunami of artificial intelligence, which affects all aspects of the business?
The Latest Google Headlines on:
Safe and fair machine behavior
[google_news title=”” keyword=”Safe and fair machine behavior” num_posts=”10″ blurb_length=”0″ show_thumb=”left”]
The Latest Bing News on:
Safe and fair machine behavior
- Iowa restaurant inspection update: Moldy eggplant, ‘not-for-sale’ meaton April 26, 2024 at 8:37 am
State, city and county food inspectors have cited restaurants and stores for hundreds of food-safety violations in recent weeks, including long-expired food, moldy vegetables and meat labeled “not for ...
- Best Real Money Online Casinos 2024 - Top USA Gambling Siteson April 23, 2024 at 2:00 pm
Best real money online casinos for US players. Sign up & play with real money to win real cash. Choose your best online casino for real money today.
- The Digital Revolution in Gambling: Navigating Through Online Casinos and Sportsbookson April 23, 2024 at 6:35 am
The digital revolution has irrevocably changed numerous sectors, with the gambling industry being one of the most significantly impacted. The transition ...
- Google all at sea over rising tide of robo-spamon April 22, 2024 at 1:30 am
What if it's not AI but the algorithm to blame? Opinion It was a bold claim by the richest and most famous tech founder: bold, precise and wrong. Laughably so. Twenty years ago, Bill Gates promised to ...
- Who Are Taylor Swift’s Tortured Poets Department Songs About? The Subjects of All 31 of Her Songson April 19, 2024 at 5:21 pm
“So Long, London,” the fifth song from The Tortured Poets Department, is believed to be about Swift’s ex-boyfriend, Joe Alwyn. The song is believed to be about Swift and Alwyn’s breakup in 2023 and ...
- Best Dogecoin Casino 2024 - Doge Gambling Sites Ratedon April 19, 2024 at 11:01 am
Dogecoin, the meme-inspired cryptocurrency, has carved out a unique niche within the online gambling industry, providing a fresh perspective on digital wagering. As more ...
- UK unions publish AI bill to protect workers from 'risks and harms' of techon April 18, 2024 at 11:30 pm
A UK federation of trades unions has published a bill designed to protect workers from “the risks and harms” of AI-powered decision-making in the workplace.
- Taylor Swift’s homage to Clara Bowon April 17, 2024 at 5:06 am
They imposed rules and codes of conduct in her contracts to hem in her behavior. The demands of work were relentless ... and the personal lives of actors and actresses were fair game. Realizing that ...
- Quiet! Our Loud World Is Making Us Sickon April 16, 2024 at 9:28 am
Experts describe ways to turn down the volume, from earbuds to smartphone apps that detect harmful noise levels ...
- In Australia, ‘Cats Are Just Catastrophic’on April 16, 2024 at 7:17 am
Feral cats take a heavy toll on the world’s wildlife, especially Down Under. The solution? Smarter traps, sharpshooters, survival camp for prey species, and the “Felixer.” ...