Can AI really code? Study maps the roadblocks to autonomous software engineering
A team of researchers has mapped the challenges of AI in software development, and outlined a research agenda to move the field forward.
Professor Emeritus Barry Vercoe, a pioneering force in computer music, dies at 87
Widely known for his Synthetic Performer, Csound language, and work on the MPEG-4 audio standard, Vercoe positioned MIT as a hub for music technology through leadership roles with the Media Lab and Music and Theater Arts Section.
Making AI-generated code more accurate in any language
A new technique automatically guides an LLM toward outputs that adhere to the rules of whatever programming language or other format is being used.
High-performance computing, with much less code
The Exo 2 programming language enables reusable scheduling libraries external to compilers.
MIT researchers introduce generative AI for databases
This new tool offers an easier way for people to analyze complex tabular data.
Creating bespoke programming languages for efficient visual AI systems
Associate Professor Jonathan Ragan-Kelley optimizes how computer graphics and images are processed for the hardware of today and tomorrow.
Natural language boosts LLM performance in coding, planning, and robotics
Three neurosymbolic methods help language models find better abstractions within natural language, then use those representations to execute complex tasks.
Probabilistic AI that knows how well it’s working
It’s more important than ever for artificial intelligence to estimate how accurately it is explaining data.
An easier way to get bugs out of programming languages
Developed at MIT, D2X is a new tool that makes it easy to debug any domain-specific programming language.
Python-based compiler achieves orders-of-magnitude speedups
Codon compiles Python code to run more efficiently and effectively while allowing for customization and adaptation to various domains.
A programming language for hardware accelerators
Researchers created Exo for writing high-performance code on hardware accelerators.
Computing our climate future
To put global climate modeling at the fingertips of local decision-makers, some scientists think it’s time to rethink the system from scratch.
A new programming language for high-performance computers
With a tensor language prototype, “speed and correctness do not have to compete ... they can go together, hand-in-hand.”
A new language for quantum computing
Twist is an MIT-developed programming language that can describe and verify which pieces of data are entangled to prevent bugs in a quantum program.
Exact symbolic artificial intelligence for faster, better assessment of AI fairness
Probabilistic programming language allows for fast, error-free answers to hard AI problems, including fairness.