Sapienza, RiccardoNg, Wai KitWai KitNg2024-10-142024-10-142024-06almahttp://hdl.handle.net/10044/1/115145This thesis focuses on the control of complex laser properties to achieve different functionalities, from emission tuning to photonic computing. We first demonstrate the dynamic lasing behaviours of self-assembly colloidal random lasers by making use of the physical control of active matter. The structural flexibility introduced in the assemblies allows for novel functionalities, such as reconfigurable and cooperative actions between active lasers. Then, we control the emission from static, on-chip semiconductor lasers via different gain modulation methods, including a machine learning approach. We also harness machine learning as a spectroscopy tool to visualise the gain profiles of complex lasing modes. Finally, we extend the on-chip laser functionalities to neuromorphic computing by drawing upon the design and modelling benefits of semiconductor network lasers. A photonic nonlinear processor with record-breaking levels of accuracy is achieved by leveraging the extremely high nonlinearity of indium phosphide network lasers for computing tasks, like nonlinear waveform transformations and machine vision challenges.Creative Commons Attribution NonCommercial LicenceThesis or dissertationhttps://doi.org/10.25560/115145