Large language models (LLMs), such as OpenAI's Codex, have demonstrated their potential to generate code from natural language descriptions across a wide range of programming tasks. Several benchmarks have recently emerged to evaluate the ability of LLMs to generate functionally correct code from natural language intent with respect to a set of hidden test cases. This has enabled the research community to identify significant and reproducible advancements in LLM capabilities. However, there is currently a lack of benchmark datasets for assessing the ability of LLMs to generate functionally correct code edits based on natural language descriptions of intended changes. This paper aims to address this gap by motivating the problem NL2Fix of translating natural language descriptions of code changes (namely bug fixes described in Issue reports in repositories) into correct code fixes. To this end, we introduce Defects4J-NL2Fix, a dataset of 283 Java programs from the popular Defects4J dataset augmented with high-level descriptions of bug fixes, and empirically evaluate the performance of several state-of-the-art LLMs for the this task. Results show that these LLMS together are capable of generating plausible fixes for 64.6% of the bugs, and the best LLM-based technique can achieve up to 21.20% top-1 and 35.68% top-5 accuracy on this benchmark.

Artificial intelligence is one of the dominant topics about the future in travel, and we're all over it at Skift. In November, OpenAI publicly released breakthrough generative AI technology, and a number of big-name travel companies have already responded. Expedia, Kayak, and more -- including multiple startups -- have started releasing experimental technologies that could lead to transformations in the way users plan and book travel. But the potential does not stop there. Advancements in AI could change the way hotels manage revenue and customer service, the way travel tech companies operate internally, and even the way airplanes and airports get designed.

"My views are my own, as are my biases." That's how Leslie Miley, investor, ex-Googler, and former CTO of the Obama Foundation, kicked off his QCon London keynote. But can the same be said for generative artificial intelligence (AI)? Not likely, as collective biases are baked in at scale, influencing everyone's views. If it keeps going unchecked, it will have devastating effects both on the Earth and the people living on it.

ChatGPT, OpenAI's text-generating AI chatbot, has taken the world by storm. It's able to write essays, code and more given short text prompts, hyper-charging productivity. But it also has a more…nefarious side. In any case, AI tools are not going away -- and indeed has expanded dramatically since its launch just a few months ago. Major brands are experimenting with it, using the AI to generate ad and marketing copy, for example.

Join top executives in San Francisco on July 11-12, to hear how leaders are integrating and optimizing AI investments for success. The sprint to develop LinkedIn's recently released generative AI tools took only three months, Ya Xu, VP of engineering and head of data and artificial intelligence (AI), told VentureBeat in an interview. The timeline, she said, was "unprecedented" for a large company like LinkedIn, given the many changes engineering and product teams implemented based on OpenAI's latest GPT models, including ChatGPT and GPT-4, as well as some open-source models. These include generative AI-powered collaborative articles, job descriptions and personalized writing suggestions for LinkedIn profiles. For example, she explained, her teams were able in just one month to generate job descriptions automatically and serve live traffic.

This has led to impressive results like producing Bach chorals,[ reference-5][ reference-6] polyphonic music with multiple instruments,[ reference-7][ reference-8][ reference-9] as well as minute long musical pieces.[ But symbolic generators have limitations--they cannot capture human voices or many of the more subtle timbres, dynamics, and expressivity that are essential to music. A different approach[ footnote-approach] is to model music directly as raw audio.[ For comparison, GPT-2 had 1,000 timesteps and OpenAI Five took tens of thousands of timesteps per game. Thus, to learn the high level semantics of music, a model would have to deal with extremely long-range dependencies.

You may have noticed some impressive video memes made with AI in recent weeks. Harry Potter reimagined as a Balenciaga commercial and nightmarish footage of Will Smith eating spaghetti both recently went viral. They highlight how quickly AI's ability to create video is advancing, as well as how problematic some uses of the technology may be. These videos remind me of the moment AI image-making tools became widespread last year, when programs like Craiyon (formerly known as DALL-E Mini) let anyone conjure up recognizable, if crude and often surreal, images, such as surveillance footage of babies robbing a gas station, Darth Vadar courtroom sketches, and Elon Musk eating crayons. Craiyon was an open source knockoff of the then carefully restricted DALL-E 2 image generator from OpenAI, the company behind ChatGPT. The tool was the first to show AI's ability to take a text prompt and turn it into what looked like real photos and human-drawn illustrations.

Los Angeles Unified, the second-largest school district in the US, immediately blocked access to OpenAI's website from its schools' network. By January, school districts across the English-speaking world had started banning the software, from Washington, New York, Alabama, and Virginia in the United States to Queensland and New South Wales in Australia. Several leading universities in the UK, including Imperial College London and the University of Cambridge, issued statements that warned students against using ChatGPT to cheat. "While the tool may be able to provide quick and easy answers to questions, it does not build critical-thinking and problem-solving skills, which are essential for academic and lifelong success," Jenna Lyle, a spokeswoman for the New York City Department of Education, told the Washington Post in early January. This initial panic from the education sector was understandable.

OpenAI on Thursday did not immediately respond to a request for comment sent overnight. In an earlier statement in response to the chatbot's false claims about the law professor, OpenAI spokesperson Niko Felix said: "When users sign up for ChatGPT, we strive to be as transparent as possible that it may not always generate accurate answers. Improving factual accuracy is a significant focus for us, and we are making progress."

We show how to use unlabeled data and a deep belief net (DBN) to learn a good covariance kernel for a Gaussian process. We first learn a deep generative model of the unlabeled data using the fast, greedy algorithm introduced by Hinton et.al. If the data is high-dimensional and highly-structured, a Gaussian kernel applied to the top layer of features in the DBN works much better than a similar kernel applied to the raw input. Performance at both regression and classification can then be further improved by using backpropagation through the DBN to discriminatively fine-tune the covariance kernel.