In nature, organisms adapt to environmental

changes through training to learn new abilities, offering

valuable insights for developing intelligent materials.

However, replicating this “adaptive learning” in synthetic

materials presents a significant challenge. This

study introduces a feasible approach to train liquid

crystal elastomers (LCEs) by integrating a mechanophore

tetraarylsuccinonitrile into their main chain,

addressing the challenge of enabling synthetic materials

to exchange substances with their environment. Inspired

by biological training, the LCEs can self-strengthen and

acquire new functionalities through mechanical stressinduced

radical polymerization. The research not only

enhances the mechanical performance of LCEs, but also

endows them with the ability to learn properties such as

flexibility, light responsiveness, and fluorescence. These

advancements are crucial for overcoming the limitations

of current materials, paving the way for the creation of

advanced intelligent soft materials with autonomous

self-improvement, akin to the adaptive skills of living

organisms.