The impact of its paper could be potential manifold but is yet to be identified.

When Apple disclosed that it would publish its research paper on artificial intelligence the pertinent question on people’s mind was would it include the data publicly? Or would it keep potential trade secrets close to its chest?

Now that Apple’s researchers have published their first ever AI research paper, their findings could be useful for computer vision technology.

Their paper focuses on developing algorithms which teach the AI to recognize objects using simulated images which is easier than from learning from photos tagged by us humans, which however suffer from issues including adapting to real-world scenarios.

The trick, however, according to Apple, is to use the popular technique of pitting neural networks against each other, with one neural network training itself to improve the realism of simulated images until they are realistic enough to match or fool a rival “discriminator” network.

However, as you may imagine, the pre-training is crucial and involves massive amount of data and time. Unfortunately despite such investments, the neural network could end up in hard-to-predict scenarios which don’t always turn up in photos.

The impact of this AI research paper is not very apparent. It could take years before its clear as to how transparent is Apple with its findings.

Nonetheless, this is a big step given the fact that its research has been on AI, an increasingly competitive field.

Unfortunately, Apple’s past reluctance to significantly contribute to scientific knowledge may keep away potential hires who want their discoveries recognised.

However, numerous papers such as these could make Apple attractive for new talent. After all Apple needs AI research for Siri, its self-driving car platform and its other AI-based projects.