Stanford’s 2025 AI Index is Out (with loads of insights)

“Stop blinking, girl. You are not being attacked by a lion. Fo–cus. It’s o…K… OMG! STOP THE MACHINES! THIS IS NOT A DRILL! HUMANITY 'S DOOMED! THAT’S IT! DISHONOUR ON YOU, DISHONOUR ON YOUR COW…” 

In retrospect, this rather accurately captures the state of my mind upon realising that I would HAVE to start using A.I. Not because I’m not into innovation– on the contrary! In fact, every so often I try a new technique, tool, technology… 

… but the truth is that I’ve always been one of those devotees of literature who sighs in awe when struck by a beautifully joined line of verse, cuddled or punched by soft or harsh literary Is, and who hoards little notebooks with translations of feelings and fleeting moments. So, naturally, I was– and remain– worried about copyright, labour, ethical and neurological implications for humankind.

Fast forward three years from the debut of the most prominent A.I. companies, they are now ubiquitous and, I must add, as good as gold for mechanical tasks such as generating tables and organizing sprawling datasets.

Stanford’s 2025 AI Index Report has just been released and shows increasing figures concerning the adoption of A.I. models in Academia and industry, Chinese and American models now running neck to neck in terms of performance, a staggering improvement of Natural Language Processing (NLP)... and, much to my surprise, Brazil showing an uptick in A.I.-related hiring, according to data from LinkedIn. 27% of generative models, the report goes on to pinpoint, are used for marketing initiatives, followed by a 19% rate of “knowledge management”.

All of which brings me to my central point: educators can no longer afford to avert their gaze. As the report itself states:

 “it is necessary to reimagine an educational program where AI competencies, inclusive of building a lens interrogating the ethics of AI in addition to technical creation, are seen as core to preparing students for a technology-powered future.” (p. 394)

The document also pinpoints recent dataset releases, A.I. uses in medicine (!), governance responses to those trends, and shifts in employability trends. It is, in every respect, well worth reading: https://hai.stanford.edu/ai-index/2025-ai-index-report