Data for 1250 articles across 250 journals (five articles per journal) were collected from March 14 to Sept 30, 2023. Articles examined were published between Feb 2, 2021, and April 25, 2023. The most common alt-text practice observed across journals was replication of figure position, such as by listing “Figure 1” or “Figure 2” (150 [60·0%] of 250 journals). It was also common for alt-text to be absent (37 [14·8%] journals) or to contain no meaningful information, such as “Figure” or “Image” (24 [9·6%] journals). In some cases, alt-text replicated the figure title (14 [5·6%] journals) or figure caption (10 [4·0%] journals). Results were similar across clinical, non-clinical, and ophthalmology journals (table). We also noted variation in alt-text practices across journals managed by the same publisher. For example, the publisher with the second highest number of journals in our sample, Elsevier, published 39 journals with alt-text that replicated figure position and four journals that provided no alt-text. We did not observe variability in alt-text practice across different articles within the same journal. Our sample contained three (1·2%) journals that provided alt-text with some context and interpretation without serving the equivalent purpose of the non-text content (ie, limited degree of interpretation), all published by the US Centers for Disease Control and Prevention (CDC).

Other times, however, Copilot is clearly just regurgitating irrelevant code samples that aren’t at all useful. Sometimes, it’s so far off base its suggestions are hilarious. (It regularly suggests that I start my components with about 25 nested divs, for example.)

But if we’re giving one of the world’s major corporations our money, in exchange for this tool that’s supposed to make us better…shouldn’t it be held to some standard of quality? Shouldn’t the results I get from a paid service at least be better than a bad StackOverflow suggestion that got down-voted to the bottom of the page (and which would probably come with additional comments and suggestions letting me know why it was ranked lower)?

As more and more of the internet is generated by LLMs, more and more of it will reinforce biases. Then more and more LLMs will consume that biased content, use it for their own training, and the cycle will accelerate exponentially.

But when I’ve asked blind colleagues what they want from Vision Pro, the conversation almost always moves to the future—not the opportunity to do computing tasks on a head-mounted device or even to watch a movie in the headset. Many blind people want Vision Pro to be an eyesight alternative or assistant, a way to see the world, identifying both the wondrous and the mundane.

It is important to make users aware of required fields upfront. This should prevent them from making submission errors and having to backtrack through a form to fix such errors. But what is the best and most accessible way to indicate required fields? This article aims to explain exactly what’s required.

This is a well-established and widely understood convention but it’s still good practice to include a note at the beginning of the form to explain that asterisks denote required fields.

This should make it clear why there is a reason that the virtual buttons in our graphical user interface should indeed look like buttons: They should commu­nicate that they can be used. When they just look like icons, they don’t do that.

In touch interfaces however, that is often not the case. There, the actual outline of the graphic – let’s say the minus – is often the only thing that can be touched.