Maybe in that scenario, it would be the author who writes that code who would be protected but the person who executes the deadly code who is breaking the law?
Holmes's famous phrase means that not all forms of speech are protected. For example, the First Amendment does not protect obscenity, child pornography, true threats, fighting words, incitement to imminent lawless action, criminal solicitation or defamation.[1]
[1]: http://1forall.us/teach-the-first-amendment/the-first-amendm...
The ELI5 of free speech is: you can say whatever you want so long as it doesn't come at a cost to another legal entity (normally people, but not always).
One example that I found very demonstrative while trying to intuitively understand FOS was a very extreme one[1]. It's a difficult but worthwhile read. The best way to understand FOS is to read up[2] on how it has been applied.
[1]: https://en.wikipedia.org/wiki/National_Socialist_Party_of_Am... [2]: https://en.wikipedia.org/wiki/Category:United_States_Free_Sp...
So while not all speech is protected, the notion of what is protected can change over time -- hopefully in a pro-liberty direction.
It should. Code is speech, so banning private speech is to create a thought-crime. Unless the author of that code intends it to be used as a weapon and facilitates that use, yes, they ought to be protected. Your hypothetical also probably isn't all that hypothetical. Pacemakers have vulns, and it isn't illegal to create a PoC proving the vuln is real. There are probably real world examples.
Yes, that's pretty much true, though actually executing that code to perform a nefarious task (or otherwise conspiring to execute the task or encouraging people to that end) may still be illegal.
How is that not the only sensible outcome?
Telling people about vulnerabilities is the only way they can defend against them. Defenders need actual exploit code to test their countermeasures against.
The person who uses the exploit to kill someone is the person who uses the exploit to kill someone.
Not a lawyer, but I don't think so. You can't just say anything you want, either (yelling "fire!" in a crowded room, slander, inciting violence, etc).
This is simply not true. Read Trope Two here [0] for a brief overview (although the entire essay is well worth reading), and [1] for a in-depth analysis of the trope.
[0] https://popehat.com/2015/05/19/how-to-spot-and-critique-cens...
[1] https://popehat.com/2012/09/19/three-generations-of-a-hackne...
The point is that the boundaries of where speech loses its First Amendment protections are very clearly defined. Now that you are aware that the "Fire!" example is not an example of unprotected speech, you would do a grave disservice to discourse if you continue to use it as an example of unprotected speech.
Of course in this case the person is a corporation; if this defense works I wonder if there will be calls for renewed scrutiny of corporate personhood. This may have been discussed in the article, but I was unable to read very far because I have an ad blocker turned on.