The tech industry is stuck on the patent issues of AI algorithms • The record

[ad_1]

The question of whether AI-generated results can be patented is affecting how tech companies protect their intellectual property.

Some of the most advanced AI technologies are systems that can produce amazing creative results. Obscure poems, short stories, and stunning digital art have all been created by machines. The human effort required to initiate these processes is often trivial: a few clicks or the typing of a text statement can lead the machine to produce something useful.

Similar generative AI models are also being applied in scientific and technological applications. Machine learning algorithms can, for example, spit out molecular compounds in efforts to discover new drugs, generate blueprints for novel chip designs, and even write code.

Under current US law, intellectual property is recognized and protected only if it is created by “natural persons”. Humans build these models, but after training, their output is often generated automatically with little help. Which begs the question, should the person who develops the AI ​​system be considered the inventor, or should the machine claim the credit? The question is a legal puzzle: Is it permissible to patent chemical compounds or software created by these systems?

Laura Sheridan, Google’s senior patent counsel, asked the US Patent and Trademark Office for guidance on the matter. Google said it has filed several patents that describe the machine learning technique used internally to automatically design and mold the components in the company’s custom AI Accelerator TPU chips used in its servers.

“We have filed patent applications on novel machine learning models. We have not pursued patents on the floor plans themselves generated by the models,” she said at the inaugural meeting of the AI ​​and Emerging Technologies (ET) Partnership Series. Held by the US Patent and Trademark Office.

Uncertainty about whether it is possible to obtain a patent protecting algorithmically generated IP, or how best to apply for it, can sometimes hinder new products, especially in the pharmaceutical and biotech industries.

Companies that rely on using AI software to create new drugs or antibodies, for example, often need to secure patents before starting clinical trials. The approval platform is critical to getting new treatments and drugs to market for patients.

“We at Google are certainly thinking a lot about the question of innovation… we’re thinking about the contribution of innovation in the AI ​​development process,” Sheridan concluded.

“It’s an area we’ve thought about a lot and I think it’s going to be very useful [USPTO] To provide guidance to patent applicants to assist with these evaluations only. They are very complicated.” ®

[ad_2]

Source link

Leave a Reply

Your email address will not be published. Required fields are marked *