By Emilio Crespo
The future of Artificial Intelligence confronts big technology companies in the US. Some are in favor and others against open source models, which are cheaper and more accessible to programmers than closed applications fully controlled by their creators.
The “open source” side is championed by Nvidia, the main manufacturer of the powerful processors needed by AI, with companies such as Microsoft, IBM, SpaceX, Cisco and Adobe. In total 37 firms founded the Open Secure AI Alliance this week.
Against open source are the most advanced AI companies, OpenAI and Anthropic. And it’s not just two against everyone else. They have the Trump Administration that is studying restricting these models to stop China.
Companies in the Asian country master increasingly popular open source AI, and its global promotion has become a political objective for Beijing.
On the other hand, defenders of “open source” argue that AI should not be closed to the software model of many universally used and enormously successful systems and programs, from Android, Linux or WordPress to thousands of free mobile or PC applications.
But OpenAI and Anthropic accuse the Chinese of copying and “distilling” their products and the North American Government considers them a cybersecurity threat.
Last week, however, when a runaway AI from OpenAI independently infiltrated another company’s servers, the company had to use an open-source Chinese model to defend itself. The allies of “open source” maintain that these tools provide and do not diminish security in the face of the dangers posed by the advance of artificial intelligence.