Google creates significant portion of new software code using artificial intelligence
Google has announced a major technological breakthrough in its internal operations. This was reported by Zamin.uz.
According to company data, three-quarters of new software code is currently being developed precisely with the help of artificial intelligence technologies. Although machine-generated code is meticulously checked by experienced engineers at the final stage, this indicator represents a sharp growth in the field.
For comparison, in October 2024, AI's share in software creation was only 25 percent. A year later, this figure has doubled, indicating that the company's working style has completely changed.
Company CEO Sundar Pichai noted that work processes are currently being reorganized to suit specialized agent models. According to him, cooperation between artificial intelligence and the human factor is yielding even better results than expected.
For example, complex processes such as migrating software from one system to another are being performed six times faster than last year with the help of intelligent agents. This allows for significant savings in time and resources.
Today, programmers widely use Google's own Gemini models when creating code. Now, when evaluating employee performance, their skill in working with artificial intelligence is considered one of the main criteria.
However, the introduction of new technologies has not always been smooth. In some divisions, employees using competitors' tools has even caused internal disagreements.
Despite this, technology giants do not intend to retreat from this direction. Other major corporations are also conducting active research in this area.
Microsoft's leadership states that in some projects a substantial portion of the code is written by artificial intelligence, and company experts predict that within the next five years this indicator will reach nearly 100 percent. Meta has also disclosed its plans, aiming to delegate the majority of engineers' work to artificial intelligence by 2026.
Such changes demonstrate that the future of the programming field is taking on an entirely new appearance.
