Can artificial intelligence make health care more efficient? (economist.com) Artificial intelligence is poised to enhance efficiency in the healthcare sector, offering solutions like AI-backed voice transcription for administrative tasks and predictive diagnostics, which can potentially save billions in healthcare expenditures. Despite challenges in integrating new technologies into established systems, AI’s potential to decentralize care and improve patient management through innovations such as virtual wards and AI-driven command centers signals a transformative shift in healthcare delivery.

How Ukraine is using AI to fight Russia (economist.com) In Ukraine’s ongoing conflict with Russia, AI has become a pivotal tool, enhancing the strategic capabilities of Ukrainian forces through various high-tech applications. This includes using AI to analyze social media and other data to impact Russian morale and logistics, optimize target acquisition with drones and digital intelligence, and bolster counter-intelligence efforts to catch spies and sanction violators, reflecting a sophisticated integration of technology into military operations.

AI models can improve corner-kick tactics (economist.com) Researchers at Google DeepMind have collaborated with Liverpool Football Club to enhance corner-kick strategies using a graph neural network (GNN), which analyses extensive player data from thousands of corner kicks to predict player movements and suggest tactical improvements. This AI model successfully predicts which player will first contact the ball with accuracy comparable to human experts and generates tactical recommendations that have been favorably reviewed by professional coaches, signaling a significant potential for AI in refining football tactics and training.

Two experts predict AI will transform companies’ understanding of themselves (economist.com) Recent advancements in AI, particularly large language models (LLMs), are revolutionizing how companies understand and improve their corporate culture. These AI tools enable leaders to extract nuanced insights from employee feedback expressed in their own words across various platforms, moving beyond the constraints of traditional engagement surveys. This shift allows for a more detailed analysis of corporate values and the identification of cultural issues, offering actionable insights that can lead to a healthier workplace environment and potentially enhance overall company performance.

What tennis reveals about AI’s impact on human behaviour (economist.com) The introduction of the Hawk-Eye ball-tracking system in tennis has significantly altered umpires’ decision-making, reducing human error in officiating by 8% since its implementation. The system, which provides a three-dimensional representation of the ball’s trajectory visible to all, has led officials to avoid potentially disruptive incorrect calls, preferring errors that are less likely to be contested by players and spectators, thus indicating a strategic shift in human judgment under AI oversight.

AI could accelerate scientific fraud as well as progress (economist.com). AI presents both opportunities and risks in scientific endeavors, as it can facilitate progress by analyzing data and generating hypotheses, but it also enables fraud through manipulation of text and images. While AI’s potential for aiding research is promising, concerns arise regarding its misuse, including the production of misleading academic articles, the pollution of data with machine-generated content, and the potential limitations of models trained on their own outputs, highlighting the need for rigorous scrutiny and accountability in scientific practices.

Authors are collaborating with AI—and each other (economist.com) “Fourteen Days,” a collaborative novel edited by Margaret Atwood and Douglas Preston, illustrates a growing trend in literature where authors work together, including notable figures like James Shapiro, Emma Donoghue, and Dave Eggers, alongside AI services like ChatGPT. Despite historical precedents of collective authorship, challenges such as reconciling diverse perspectives and the acceptance of AI in literary circles persist, highlighting the ongoing dominance of individual authors in creative endeavors.

Why AI needs to learn new languages (economist.com) Efforts to enhance the multilingual capabilities of artificial intelligence, spurred by disparities in language performance, particularly in low-resource languages, are underway globally, with approaches including optimizing tokenizers, improving training datasets, and refining models post-training, aimed at democratizing AI access and effectiveness across diverse linguistic landscapes, albeit with challenges such as illiteracy rates and the potential dominance of major tech companies looming.

AI models will become smaller and faster (economist.com) In 2024, advancements in artificial intelligence will focus on making AI models smaller, faster, and more efficient, with improvements in size, data quality, and applications, paving the way for new techniques such as retrieval augmented generation and fine-tuning to enhance AI’s capabilities across various fields. Despite the rapid progress, the search for artificial general intelligence continues, with researchers acknowledging that current AI models may be a step in the journey but not the final solution.

The dawn of the omnistar (economist.com) In 2023, the rise of artificial intelligence (AI) is shaking up the entertainment industry, with AI-generated content impacting everything from music charts to Hollywood stars, prompting concerns over copyright and the future of fame. Despite fears of diluting star power, AI is poised to amplify the influence of mega-celebrities, enabling them to be omnipresent across various formats and markets, while also presenting challenges such as legal uncertainties and potential audience boredom.