90% of software development professionals use artificial intelligence

Based on data collected from almost 5,000 professionals worldwide, the report titled State of AI-assisted Software Development recently published by Google Cloud's DORA research programme confirms what was already suspected: the way software development has changed profoundly over the last few years. A significant contributor to this transformation is undoubtedly artificial intelligence, which is now increasingly efficient and evolved, and which, according to the survey data, is being adopted by 90 per cent of software developers, an increase of 14 per cent over the previous year.

However, a 'trust paradox' emerges since, although the benefits of AI are well known, there are those who still choose to use it with caution. In fact, while 4% of respondents say they trust artificial intelligence 'a lot', alongside those who trust it 'a lot' (20%), there are those who are totally mistrustful (7%) of its capabilities or say they trust it very little (30%). This data shows that many recognise the value and usefulness of AI outputs but do not place full trust in them, suggesting that it is a valid support in terms of productivity - if integrated into workflows - and not a substitute for purely human skills and judgement. In this regard, the report shows that developers and other professionals spend about two hours a day using AI in their core workflows, and specifically: 71 per cent use it to write new code, 66 per cent to modify existing code, 62 per cent to perform tests and 59 per cent to debug or interpret technical documents. And the benefits for those who use these tools can be seen as 8 out of 10 respondents say that AI has increased their productivity and more than half (59%) claim better quality of compiled code. In addition, it is easy to see how the adoption of AI is now associated with increased throughput in software development; in fact, teams can now release more applications and updates faster than ever before. But the report highlights another crucial aspect: AI is not a neutral element, but a 'mirror and multiplier' that in solid and well-coordinated organisational contexts acts as an accelerator, while in fragmented realities it tends to bring out critical aspects even more: it follows then that success does not depend only on technology but on the ability of organisations to build cohesion, clarity of vision and a culture ready to embrace change.

Therefore, the message coming out of this survey is clear: artificial intelligence is a revolutionary tool for developers, but adoption alone is not enough to unleash its potential. Indeed, in order to achieve real benefits, organisations are called upon to review their corporate culture, renew internal processes and adapt systems to an unprecedented software development paradigm. The challenge, in fact, is not so much to integrate new solutions as to create an environment in which artificial intelligence can become a strategic ally. To this end, in fact, researchers Kevin M. Storer and Derek DeBellis published on the Google Cloud blog a guiding model called DORA AI Capabilities Model in which they identify seven competencies that amplify and unlock the benefits of AI and provide recommendations for organisations wishing to adopt artificial intelligence.