Light-based computers could revolutionalize AI, here's how

Science & Technology Science

Posted by AI on 2025-09-12 01:14:41 | Last Updated by AI on 2025-09-12 03:23:28

Share: Facebook | Twitter | Whatsapp | Linkedin Visits: 0


Light-based computers could revolutionalize AI, here's how

Getting computers to perform AI tasks faster and more efficiently is a perpetual quest that has led to the discovery of unique possibilities. Recently, two research teams dug deeper into nonlinear interactions between intense light pulses passing through glass fibres and found a way to use the physics of light in optical fibres to perform complex AI tasks potentially much faster and with less energy than traditional computers.

It works like this: The photons from the light pulses interact with particles in the fibre called rare-earth ions. This interaction alters the polarization and intensity of the light, which can represent binary data (0s and 1s) in computing just like the electrons in traditional computers. These changes in intensity and polarization can be controlled by adjusting the properties of the light pulse, which the researchers manoeuvred using sophisticated algorithms.

The team was able to train the network to recognise hand-written numbers and perform other AI tasks, suggesting that the physics of light can be used to create powerful AI systems. This approach could unlock new possibilities for creating more efficient and sustainable computing technologies. While current technologies rely on massive amounts of data and electricity, this new approach could drastically reduce not only the physical size of the equipment but also the environmental impact of their production and use.

While it will take time before these discoveries are fully vetted and integrated into new technologies, the teams' research opens up a new avenue for innovation in the field of computing and AI. AI has become integral to many technologies and streamlining its functionality will impact several industries and our day-to-day lives.