New research conducted by Modl.ai, a company specialized in QA (i.e. Quality Assurance) through artificial intelligence, revealed that 77% of video games developers do not perform adequate tests before the publication of their qualifications. On a sample of almost 300 professional developers in the United States, it highlights a reality now under everyone’s eyes: the launch of games in far from optimal conditions has become the standard. 23% of developers do you think they invest enough during the QA. \ r \ nil main problem? The budgets intended for tests do not grow hand in hand with the growing complexity of modern games. \ R \ nun since it sounds almost ironic, considering that today video games cost astronomical figures to be developed. A disaster. Cyberpunk 2077 (which you find on Amazon) took years to become a product at the height of the promises. Even Starfield, a project on which Bethesda has focused a lot, was launched with more problems than certainties. \ Nper tackle the problem, most developers look to artificial intelligence as possible salvation. However, the transition is far from imminent: at the moment only 18% of companies are really investing in the integration of this technology. The usual justifications: the market has changed, the development is more complex, the expectations are higher (and there are already those who put their hands forward). \ r \ Nor, the more budgets increase, the more the games arrive on the market in pitiful condition. Perhaps it is appropriate to say that, in addition to the technical difficulties, the problem is another: the desire to publish finished and functioning games seems to have become secondary compared to the race at day-one.