study

The models tended to answer questions inconsistently, which reflects biases embedded in the data used to train the models.

Study finds that AI models hold opposing views on controversial topics

Data from various surveys show that there’s an exodus in academia where it concerns top AI talent — and where they end up.

There’s an AI ‘brain drain’ in academia

Will AI automate human jobs, and — if so — which jobs and when? That’s the trio of questions a new research study from MIT’s Computer Science and Artificial Intelligence…

New MIT CSAIL study suggests that AI won’t steal as many jobs as expected

Most humans learn the skill of deceiving other humans. So can AI models learn the same? Yes, the answer seems — and terrifyingly, they’re exceptionally good at it. A recent…

Anthropic researchers find that AI models can be trained to deceive

Today during its first-ever dev conference, OpenAI released new details of a version of GPT-4, the company’s flagship text-generating AI model, that can understand the context of images as well…

As OpenAI’s multimodal API launches broadly, research shows it’s still flawed

The Frontier Model Forum, an industry body focused on studying “frontier” AI models along the lines of GPT-4 and ChatGPT, today announced that it’ll pledge $10 million toward a new…

AI titans throw a (tiny) bone to AI safety researchers

Sometimes, following instructions too precisely can land you in hot water — if you’re a large language model, that is. That’s the conclusion reached by a new, Microsoft-affiliated scientific paper…

Microsoft-affiliated research finds flaws in GPT-4

It’s no secret that OpenAI’s viral AI-powered chatbot, ChatGPT, can be prompted to say sexist, racist and pretty vile things. But now, researchers have discovered how to consistently get the…

Researchers discover a way to make ChatGPT consistently toxic

Early in the pandemic, a number of researchers, startups and institutions developed AI systems that they claimed could diagnose COVID-19 from the sound of a person’s cough. At the time,…

AI does a poor job of diagnosing COVID-19 from coughs, study finds

A recent study finds that software engineers who use code-generating AI systems are more likely to cause security vulnerabilities in the apps they develop. The paper, co-authored by a team…

Code-generating AI can introduce security vulnerabilities, study finds

Code-generating systems like DeepMind’s AlphaCode, Amazon’s CodeWhisperer and OpenAI’s Codex, which powers GitHub’s Copilot service, provide a tantalizing look at what’s possible with AI today within the realm of computer…

Hugging Face and ServiceNow launch BigCode, a project to open source code-generating AI systems

Earlier this year, Twitter launched the Twitter Moderation Research Consortium (TMRC), a group of experts from across academia, civil society, nongovernmental organizations and journalism dedicated to studying Twitter’s platform governance…

Twitter allows more researchers to access platform data

Research in the field of machine learning and AI, now a key technology in practically every industry and company, is far too voluminous for anyone to read it all. This…

Perceptron: AI that solves math problems, translates 200 languages and draws kangaroos

Cohere, a startup creating large language models to rival those from OpenAI, today announced the launch of a nonprofit research lab: Cohere For AI. Headed by Google alum Sara Hooker,…

AI startup Cohere launches a nonprofit research lab

Research in the field of machine learning and AI, now a key technology in practically every industry and company, is far too voluminous for anyone to read it all. This…

Perceptron: Robots that feel pain and AI that predicts soccer players’ movements

Research in the field of machine learning and AI, now a key technology in practically every industry and company, is far too voluminous for anyone to read it all. This…

Perceptron: Risky teleoperation, Rocket League simulation and zoologist multiplication

Timed to align with Build 2022, Microsoft today open sourced tools and datasets designed to audit AI-powered content moderation systems and automatically write tests highlighting potential bugs in AI models.…

Microsoft claims its new tools make language models safer to use

Research in the field of machine learning and AI, now a key technology in practically every industry and company, is far too voluminous for anyone to read it all. This…

Deep Science: AI cuts, flows and goes green

European governments have been bringing the hammer down on tech in recent months, slapping record fines and stiff regulations on the largest imports out of Silicon Valley. Despite pleas from…

Tech regulation in Europe will only get tougher

Despite well-documented concerns over an economic slowdown in China, the country’s luxury goods market is still seeing opulent growth according to a new study. Behind secular and demographic tailwinds, the…

Slowdown or not, China’s luxury goods still seeing high-end growth

Featured Article

Pew: A majority of U.S. teens are bullied online

A majority of U.S. teens have been subject to online abuse, according to a new study from Pew Research Center, out this morning. Specifically, that means they’ve experienced at least one of a half-dozen types of online cyberbullying, including name-calling, being subject to false rumors, receiving explicit images they didn’t…

Pew: A majority of U.S. teens are bullied online

Kelly Dillon, a Doctoral Candidate at the Ohio State University School of Communication, found that people generally won’t act when they see someone bullied online. This news, while not surprising…

A Chat With The Researcher Who Found Online Bystanders Won’t Stop Bullying

[This post has been updated to include comments from Adam Kramer, the Facebook employee who coauthored the study, and new information from Cornell indicating that the study was funded internally. ]…

Facebook And The Ethics Of User Manipulation [Updated]

A fascinating new study explores, in exquisite detail, the hidden secrets of innovation. Rampant copycats, the authors find, are a surprising key to discovering creative solutions. Imitators must, however, be…

Study: Copycatting, Diverse Teams, And Transparency Are Keys To Innovation

If you’ve ever wondered why birthers, inside jobbers, and other conspiracy theorists won’t simply take the truth at face value, two researchers at OSU have found that we tend to…

Study Finds That We Still Believe Untruths Even After Instant Online Corrections

Much of the time when you talk about browser market share, it’s in the context of IE vs. Chrome vs. all the rest. But new data released from e-commerce technology…

Ecommerce Study Finds Mobile Safari To Be Fastest Growing Web Browser

Juniper Research released a report today that found that global mobile network operator revenues are set to exceed $1 trillion by 2016. This would be something to celebrate were it…

Wireless provider Mobistar recently announced the results of their survey which sought out cell phone usage by gender. If stereotypes are to be believed, then the results are hardly shocking.…

There’s a new study out that says having tax dollars go towards free cell phones for the poor actually creates money for the economy. They’ve been dubbed “Obama Phones” in…

This is just what Alone Together predicted: people, particularly young people, feeling more comfortable with themselves online than they are offline. Black is white, up is down, and the Mets…