Artificial intelligence is facilitating human life in many aspects. Previous artificial intelligence has been mainly focused on computer algorithms (e.g. deep-learning and extremelearning) and integrated circuits. Recently, all-optical diffractive deep neural networks (D 2 NN) were realized by using passive structures, which can perform complicated functions designed by computer-based neural networks at the light speed. However, once a passive D 2 NN architecture is fabricated, its function will be fixed. Here, we propose a programmable artificial intelligence machine (PAIM) that can execute various intellectual tasks by realizing hierarchical connections of brain neurons via a multi-layer digital-coding metasurface array. Integrated with two amplifier chips in each meta-atom, its transmission coefficient covers a dynamic range of 35 dB (from -40 dB to -5 dB), which is the basis to construct the reprogrammable physical layers of D 2 NN, in which the digital meta-atoms make the artificial neurons alive. We experimentally show that PAIM can handle various deep-learning tasks for wave sensing, including image classifications, mobile communication coder-decoder, and real-time multi-beam focusing. In particular, we propose a reinforcement learning algorithm for on-site learning and discrete optimization algorithm for digital coding, making PAIM have autonomous intelligence ability and perform self-learning tasks without the support of extra computer.

A digital-coding programmable metasurface (DCPM) is a type of functional system that is composed of subwavelength-scale digital coding elements with opposite phase responses. By configuring the digital coding elements, a DCPM can construct dynamic near-field image patterns in which the intensity of each pixel of the image can be dynamically and independently modulated. Thus, a DCPM can perform both spatial and temporal modulations. Here, this advantage is used to realize multichannel direct transmissions of near-field information. Three points are selected in the near-field region to form three independent channels. By applying various digital phase codes on the DCPM, independent binary digital symbols defined by amplitude codes (namely, weak and strong amplitudes) are transmitted through the three channels. The measured near-field distributions and temporal transmissions of the system agree with numerical calculations. Compared with the conventional multichannel transmission, the proposed mechanism achieves simultaneous spatial and temporal modulations by treating DCPM as an energy radiator and information modulator, thereby enduing DCPM with high potential in near-field information processing and communications.

Brain-computer interfaces (BCIs), invasive or non-invasive, have projected unparalleled vision and promise for assisting patients in need to better their interaction with the surroundings. Inspired by the BCI-based rehabilitation technologies for nerve-system impairments and amputation, we propose an electromagnetic brain-computer-metasurface (EBCM) paradigm, regulated by human’s cognition by brain signals directly and non-invasively. We experimentally show that our EBCM platform can translate human’s mind from evoked potentials of P300-based electroencephalography to digital coding information in the electromagnetic domain non-invasively, which can be further processed and transported by an information metasurface in automated and wireless fashions. Directly wireless communications of the human minds are performed between two EBCM operators with accurate text transmissions. Moreover, several other proof-of-concept mind-control schemes are presented using the same EBCM platform, exhibiting flexibly-customized capabilities of information processing and synthesis like visual-beam scanning, wave modulations, and pattern encoding.

Topological photonics has revolutionized our understanding of light propagation, providing a robust way to manipulate light. So far, most of studies in this field are focused on designing a static photonic structure. Developing a dynamic photonic topological platform to switch multiple topological functionalities at ultrafast speed is still a great challenge. Here we theoretically propose and experimentally demonstrate a reprogrammable plasmonic topological insulator, where the topological propagation route can be dynamically changed at nanosecond-level switching time, leading to an experimental demonstration of ultrafast multi-channel optical analog-digital converter. Due to the innovative use of electric switches to implement the programmability of plasmonic topological insulator, each unit cell can be encoded by dynamically controlling its digital plasmonic states while keeping its geometry and material parameters unchanged. Our reprogrammable topological plasmonic platform is fabricated by the printed circuit board technology, making it much more compatible with integrated photoelectric systems. Furthermore, due to its flexible programmability, many photonic topological functionalities can be integrated into this versatile topological platform.

Digital coding metasurface, which provides a new approach to link the physical world and information science, has been quickly developed in recent years. However, all previously reported metasurfaces cannot achieve independent controls of different polarizations in both transmission and reflection spaces at the same time. In this work, a reconfigurable anisotropic digital coding metasurface loaded with electronically controlled PIN diodes is proposed that can independently manipulate not only the near/far‐field pattern but also the transmission and reflection modes of the electromagnetic waves under different polarizations. As a validation example, a multifunctional holographic imaging metasurface is designed, fabricated, and measured. Both simulated and measured results show that orthogonally polarized waves (vertical and horizontal polarizations) can be manipulated to achieve different images, and the transmission and reflection modes of the differently‐polarized images can be independently controlled in real time by changing the state of the loaded PIN diodes.