The opinion archive provides access to past opinion stories from Communications of the ACM and other sources by date.
Perfectly safe algorithmic systems are not possible, but safer systems are.
Medical AI applications need to overcome transparency and trust issues.
Hype has been a primary driver of the excitement around LLMs.
We often judge the utility of a new technology on how well it deals with the problems of older ones.
We need to change rules and institutions while still promoting innovation to protect people from faulty artificial intelligence.
Promises that self-driving cars would revolutionize transportion continue to fall short.
The technology is simply not ready to be used like this at this scale.
A professor explains why he is allowing students to incorporate ChatGPT into their writing process instead of banning it.
Removing immunity for algorithmic recommendations would make it nearly impossible for social media platforms as we know them to function.
The Internet is a toxic waste dump.
AI is being designed and deployed by corporate America in ways that will disempower and displace workers and degrade the consumer experience.
The U.S. healthcare system is plagued by inefficiencies, in part due to deficiencies of health data interoperability between electronic health records and other health data systems.
The scientific advances announced in recent months are thrilling, capable of sustaining our economic and military competitiveness for decades to come.
Why companies must build safety into tech products.
A wave of research improves reinforcement learning algorithms by pre-training them as if they were human.
Conversational AI is a game-changer for science; here's how to respond.
Conversation without change is just a five-finger exercise.
How active-learning techniques can benefit students in computing courses.
It is time to get the POSIX elephant off our necks.
The equation Ethics + AI = Ethical AI is questionable.
Why deep learning will not replace programming.
Exploring Black faculty at computer science research departments where Ph.D. programs exist.
Proposing a community-based system for model development.
Seeking to make machine learning more dependable.