Early days for small language models and AI at the edge

Small language models could be more cost-effective to deploy than LLMs, offering greater privacy and performing specific tasks better. But is it too early for SLMs?