- The survey of 800 developers reveals that more than half believe that AI can code better than humans
- But most are positive on changes provide that AI brings
- Reservations remain, employment security, data confidentiality and inaccuracies always causing concerns
With large language models becoming more and more an inevitable characteristic in software spaces, and new clutch research has revealed a little more than half (53%) of developers think that LLM can be better coded than most humans.
As with many industries, AI tools are widely adopted in the software world, with 49% of senior developers and team leaders indicating that they use such articles every day, and 78% of developers use them several times a week.
Perhaps surprising, the study revealed an extremely favorable inclination for AI, 42% feeling positive about its use – against only 10% of respondents with concerns, and 8% who are “skeptical”.
Persistent fears
Although most developers are looking forward to using AI, the study revealed serious reserves. Mainly, developers are concerned about AI management of sensitive data, 24% citing data confidentiality as the main concern.
Not only that, because some (14%) are concerned about the movement of work, as well as inaccuracy (14%), and a loss of creativity (13%) – while only 8%are not at all concerned.
The labor market is evolving rapidly with the integration of AI, because 79% say that IA skills will soon be a requirement for developers, and 45% think that AI will reduce the barrier for junior developers in the future.
More than two -thirds (37%) believe that new coders could be replaced or over -performed by AI – and after repeated news of the technological sector layoffs, software engineers are increasingly worried that AI ruins the labor market.
The market for junior developers seems more and more difficult, with entry -level jobs disappearing thanks to LLM, leaving a potentially disastrous scenario once the developers of higher level take their retirement in which companies find themselves with very little real experience – based heavily on AI without qualified experts to supervise the work.
Even now, more than half of the developers (59%) admit to rely on the code generated by the AI that they do not fully understand. That being said, most do not see AI as a threat, 80% describing tools as catalysts – perhaps indicating a preference for short -term productivity compared to longer -term planning.
