Back when US President Donald Trump (and his challenger Joe Biden) were students, artificial intelligence (AI) had a different connotation. And it is just as well for the White House incumbent in particular, as a new study of the correlation between social media posts and examination scores using AI indicates that content with words and phrases written entirely in capital letters portends lower academic performance. Having trained on 1.9 billion words' worth of content, AI found that the use of all capital letters (–0.08), emojis (–0.06) and exclamation marks (–0.04) corresponded to lower academic performance. While Indians may be intrigued by this conclusion, it is not surprising that the study has shown content with extensive vocabularies, use of terms relating to science and processes, literary flourishes and quotations correspond to better exam scores than those that show a predilection for spelling mistakes and allusions to computer games, horoscopes and military concepts. Of course, Mr Trump would aver that examiners (and AI) are inherently biased and discriminate against those who do not conform to the usual parameters of scholarly expression.

If umifenovir, a broad-spectrum antiviral, can fight COVID-19, then computer-designed synthetic routes could make it easy and cheap to produce. Science's COVID-19 reporting is supported by the Pulitzer Center and the Heising-Simons Foundation. As scientists uncover drugs that can treat coronavirus infections, demand will almost certainly outstrip supplies--as is already happening with the antiviral remdesivir. To prevent shortages, researchers have come up with a new way to design synthetic routes to drugs now being tested in some COVID-19 clinical trials, using artificial intelligence (AI) software. The AI-planned new recipes--for 11 medicines so far--could help manufacturers produce medications whose syntheses are tightly held trade secrets.