I respectfully disagree. TCP or DNS or WiFi or other technologies are merely means to achieve some result. A tool.

Twitter, like most services, is also built using various technologies and tools. But its main distinguishing property is that it has a large number of users who, for various reasons, are interested in what some other users have to say. Creating such social connections is its main goal.

Now, one might use e.g TCP to spread hate speech all over the internet. But apart from computers dropping these packets, almost no real person will be listening.

Contrast that with a Twitter account that has ~10k followers. If the hate speech is spread from there, it can get a lot of audience very quickly.

Twitter is one of many enablers and hosts of large online communities of people. As such, it should have, in my opinion, some responsibility regarding what goes on within these communities. At a minimum, it should disallow the dissemination of hate speech, actively seek and remove it and block the users who repeatedly spread it.

That being said, it might be difficult to precisely define what constitutes a hate speech and what not. But Twitter should at least be trying.