The article discusses the challenges and opportunities in designing wireless communication systems that support short packets, which are essential for 5G and future networks. Current systems are optimized for long packets, but 5G needs to handle short packets for applications like Machine-to-Machine (M2M) communications and ultra-reliable low-latency communications (URLC). Short packets require new design principles because metadata (control information) becomes significant compared to the payload, making traditional methods suboptimal. The paper reviews recent advances in information theory that govern short-packet transmission. It applies these principles to three scenarios: two-way channels, downlink broadcast channels, and uplink random access channels. These examples show how control information can be optimized for short packets, highlighting the need for new wireless protocols. The article introduces a performance metric, the maximum coding rate at finite packet length and finite packet error probability, which is more relevant for short packets. It explains how to evaluate this metric for additive white Gaussian noise (AWGN) and fading channels, providing insights into engineering design. The paper also discusses the tradeoff between diversity, multiplexing, and channel estimation in multiple-input multiple-output (MIMO) fading channels. It shows that for small packet lengths, there is a fundamental tradeoff between using spatial diversity to reduce error probability or increasing data rate through spatial multiplexing. The analysis reveals that for certain scenarios, spatial diversity may be more beneficial than multiplexing. The article concludes that the traditional information-theoretic metrics like capacity and outage capacity are not sufficient for short-packet communications. Instead, a more refined analysis of the maximum coding rate is needed to capture the tradeoffs between reliability, throughput, and resource utilization. The paper also highlights the importance of considering channel estimation overhead in the design of wireless protocols for short packets.The article discusses the challenges and opportunities in designing wireless communication systems that support short packets, which are essential for 5G and future networks. Current systems are optimized for long packets, but 5G needs to handle short packets for applications like Machine-to-Machine (M2M) communications and ultra-reliable low-latency communications (URLC). Short packets require new design principles because metadata (control information) becomes significant compared to the payload, making traditional methods suboptimal. The paper reviews recent advances in information theory that govern short-packet transmission. It applies these principles to three scenarios: two-way channels, downlink broadcast channels, and uplink random access channels. These examples show how control information can be optimized for short packets, highlighting the need for new wireless protocols. The article introduces a performance metric, the maximum coding rate at finite packet length and finite packet error probability, which is more relevant for short packets. It explains how to evaluate this metric for additive white Gaussian noise (AWGN) and fading channels, providing insights into engineering design. The paper also discusses the tradeoff between diversity, multiplexing, and channel estimation in multiple-input multiple-output (MIMO) fading channels. It shows that for small packet lengths, there is a fundamental tradeoff between using spatial diversity to reduce error probability or increasing data rate through spatial multiplexing. The analysis reveals that for certain scenarios, spatial diversity may be more beneficial than multiplexing. The article concludes that the traditional information-theoretic metrics like capacity and outage capacity are not sufficient for short-packet communications. Instead, a more refined analysis of the maximum coding rate is needed to capture the tradeoffs between reliability, throughput, and resource utilization. The paper also highlights the importance of considering channel estimation overhead in the design of wireless protocols for short packets.