midjourney MidJourney Beta Transforms Simple Text Prompts Into Stunning AI Artwork The MidJourney beta wows users by generating highly detailed, eerie, and beautiful images directly from text descriptions. The tool runs through Discord and allows creators to easily upscale and refine their unique AI-generated artwork.
deepmind DeepMind AlphaCode Achieves Median Rank in Competitive Programming DeepMind's AlphaCode system writes computer programs at a competitive level, placing within the top 54% of participants on Codeforces. The AI system uses massive-scale code generation and smart filtering to solve novel problems.
alphacode DeepMind AlphaCode Achieves Top-Half Ranking in Competitive Programming DeepMind introduces AlphaCode, an AI system that solves complex competitive programming problems by combining large-scale language models with smart filtering techniques. The system achieves an average ranking within the top 54.3% against thousands of human participants on the Codeforces platform.
artificial intelligence DeepMind Tackles Fusion Energy as AI Research Evolves March 2022 brings major AI updates, including DeepMind's breakthrough in controlling fusion plasma and new tools like PyTorch's TorchRec for recommender systems.
mit MIT Researchers Show Synthetic Data Can Rival Real Images for AI Training MIT researchers develop a generative model that creates synthetic data to train AI for image classification, potentially saving millions in dataset costs. The method sometimes outperforms models trained on real data.
artificial intelligence AI Transforms Cancer Target Identification and Drug Discovery A new review highlights how artificial intelligence accelerates the identification of novel anticancer targets and drug candidates by analyzing complex biological networks. The study details machine learning and network-based methods that provide a quantitative framework for understanding cancer.
nlp New Survey Tackles Compression Techniques for Large Language Models A new paper accepted to AAAI 2023 reviews current methods for compressing and accelerating pretrained language models to reduce their massive energy costs and inference delays. The survey focuses specifically on the inference stage to help deploy these models on edge and mobile devices.
dynamic neural networks Dynamic Neural Networks Offer Efficient Alternative for Massive NLP Models A new survey highlights how dynamic neural networks effectively scale natural language processing models without linear increases in computing power. The research explores methods like skimming, mixture of experts, and early exit to handle trillion-parameter models.
deepmind DeepMind AlphaCode Achieves Median Human Level in Coding Competitions DeepMind's AlphaCode system writes computer programs at a competitive level, ranking in the top 54% of participants on Codeforces. This milestone marks the first time an artificial intelligence achieves competitive performance in complex programming contests.
nfl NFL Tackles Game Evolution With Advanced AI Analytics The NFL partners with AWS to process over 500 million data points per season, transforming player safety, scouting, and fan engagement through artificial intelligence.
deepmind DeepMind AlphaCode Achieves Median Human Level in Coding Competitions DeepMind's AlphaCode system writes computer programs at a competitive level, ranking in the top 54% of participants on Codeforces. The tool uses massive-scale code generation and smart filtering to solve novel problems.
instructgpt How InstructGPT Uses RLHF to Align AI with Human Preferences OpenAI's 2022 InstructGPT research introduces a three-stage training process using reinforcement learning from human feedback to make AI more helpful, honest, and harmless. This method solves the critical gap between what language models generate and what users actually want.
openai Smaller AI Models Outperform GPT-3 Through Human Feedback Training OpenAI researchers introduce InstructGPT, a language model trained with human feedback to better follow user instructions. Despite having 100 times fewer parameters, this aligned model consistently outperforms the massive GPT-3 in human evaluations.
openai OpenAI Outlines Three-Part Strategy to Align Future AI Systems OpenAI adopts an iterative, empirical approach to AI alignment, focusing on scalable training signals based on human intent. The company relies on three main pillars, including using human feedback to train models like InstructGPT.
mustafa suleyman DeepMind Co-Founder Mustafa Suleyman Now Leads Microsoft AI British AI entrepreneur Mustafa Suleyman currently serves as the CEO of Microsoft AI after co-founding DeepMind and Inflection AI. His career spans from establishing a major mental health support service as a teenager to shaping the modern artificial intelligence industry.