The Internet as a whole, including access technologies and cellular networks, has undergone a significant evolution since its original setup. However, key TCP/IP functionalities, such as multicast and service differentiation, have not been widely deployed and core protocols have not always kept up with advancements. Meanwhile, the number and diversity of online applications and content have skyrocketed, not to mention the amount of traffic they generate, while network operators encountered difficulties in diversifying their portfolio of services. In addition, business models have changed a lot and new ones have emerged. The combination of these factors has resulted in two momentous outcomes: i) the network is diverse and patchy, with performance greatly varying across different regions, whereas important functionalities are provided unevenly and/or as network overlays; ii) new players have entered the market, narrowing the role of traditional network providers: hyperscalers, providing applications and content as well as cloud services and connectivity; CDN providers; cybersecurity providers; IXPs and intermediaries, but also new connectivity providers, such as utilities and municipalities. Despite impressive improvements in network protocols and performance, the question remains whether the architecture and current service offerings of the public Internet are at par with the requests of users, and especially of innovative businesses. This talk will discuss some options for the evolution of the network. Shall we accept the current situation and confide in good/over dimensioning and market dynamics? Or should we consider innovative network architectures? Should we update net neutrality principles and introduce so called specialised services on top of basic Internet access services? In this context, we will analyse the role of stakeholders in the changing landscape and identify open challenges.