On 7/20/2022 9:41:45 PM, Spiros Bousbouras wrote:
> Dear Mr. Musk,
>
> I have read in the news that you are (or were) interested in buying
> Twitter in order to promote free speech. I applaud your goals , I care a
> lot about free speech myself. But let me suggest that the operational
> model of Twitter and free speech do not fit well together.

There are more than a handful of folks who believe that Musk's discussion
regarding purchasing Twitter had far less to do with a desire for the
preservation of Free Speech, and more to do with stock market manipulations.
Twitter's stock price has been all over the map since the announcement, Musk
leveraged Tesla stock more than his cash-on-hand, and Twitter having to disclose
the actual number of users and bots as a matter of public record through the SEC
filings altered the price advertisers were willing to pay for ad space. Whether
Twitter was an unwitting pawn in Musk restructuring his Tesla ownership or there
was some sort of feud between Musk and Twitter's execs is unclear, but a lot of
people - myself included - are not convinced that Musk's primary concern was
altruistic.

Personally, the reason I'm unconvinced that free speech was his goal was because
Musk has some pretty solid resources at his disposal, both human and
technological. I'm sure he could have gotten a thousand people together, bought
an office building (if he doesn't have a spare already), gotten a couple of racks
of servers and hard drives from Dell or HP, forked Mastodon and spun up his own
Twitter competitor whose selling point was "better terms of service, clear due
process for violations, and no ads or tracking scripts for two years"...and he
probably would have had modest success with it AND spent $43 billion less than he
offered Twitter.

> An absolute concept of free speech is that if persons A and B want to
> exchange views or information about whatever issue , no entity C (where
> "entity" means individual or corporation or government) should be able to
> prevent it. In other words , the decision should be entirely up to A and B
> and noone else.

I submit that this still exists, for the most part. E-mail is still mostly
unaffected by this, and while both Microsoft and Google are likely to hand over a
user's inbox to law enforcement whenever asked, they're unlikely to censor
contents. The protocol itself has all the functionality you describe; it's
decentralized and federated, and anyone can spin up a mail server if they so
choose. There are also a number of chat applications that handle synchronous
communication in a similar manner. Signal and Telegram have so far managed to
hold up to some scrutiny, while Rocketchat and Mattermost and Matrix allow users
to spin up their own chat servers and federate them as well.

The statement above assumes one-to-one communication, while Twitter's claim to
fame is one-to-many communication...and that's why the question arises with
Twitter in a way that it doesn't with E-mail.

> There are physical limitations which make such absolute free speech
> impossible and it may not even be desirable. But it is also very far from
> a desirable level of free speech if a single entity , like Twitter , can
> restrict people's ability to communicate with each other. Obviously
> Twitter isn't the only form of online communication but my overall point
> is that online discussion is too centralised at present , that this
> centralisation negatively affects freedom of speech and plurality of
> opinions therefore the right way forward is to give emphasis to more
> decentralised methods of online discussion.

Centralization also has its benefits, if we're going to be real about it. If it
didn't, Gmail wouldn't be the default it is today. Ever try to solve an e-mail
flow issue? User->Server->Filter->Internet- >Filter->Server->User, any one of
those links can go wrong. They're worth having for the very reasons you specify,
but we can't truly solve an issue if we're not honest about why it is chosen.

Yes, Twitter brings censorship with it, but it also brings message amplification
to it. Reddit does this as well. Though Reddit is admittedly susceptible to
groupthink, lets users upvote/downvote and sort by those votes, allowing
generally-more-desirable content to be sifted from the generally-less-desirable
content, without actually censoring anyone (in principle, anyway). As much as I
appreciate the true egalitarianism of Usenet, it is disingenuous to paint the
algorithms at Twitter (and the more human one at Reddit) as completely without
merit. Your post and some random cryptocurrency spam have two different values.
The relatively low user count of Usenet at the moment is pretty much the primary
reason why your post wasn't bordered by a thousand crypto bot spam messages and
the protocol makes it extremely difficult to solve this problem.

Retroshare, an interesting chatroom/IM/Usenet/E-mail/p2p file sharing app, has
the tech part on lock, but my time in their Usenet-esque section was rather
unnerving. Actual-antisemitism (i.e. calls to 'finish the job'), bomb-making
instructions, unhinged conspiracy theories and 'erotica' involving violence were
just a handful of the topics represented. I'm not quite sure where the line is
drawn, but "free speech for them too" meant Retroshare was philosophically
consistent at the expense of making the community somewhere I'd never recommend
to anyone else. A community that *can* become like that *will* become like that
eventually.

> And as it turns out , there is such a method. Not only that but it is one
> of the oldest forms of online dicussion : usenet !
>
> I will guess that you have already encountered usenet although perhaps not
> recently. It is a lot less popular than what it was some decades ago but
> it is still going strong. From the point of view of freedom of speech it
> has many advantages over Twitter : it is decentralised , meaning a large
> number of servers controlled by different entities instead of servers
> ultimately controlled by a single entity which can command that this or
> that should be censored. Usenet is based on open standards for which there
> exist already a large number of implementations both of servers and
> clients and also programming libraries for many programming languages. So
> whether one wants to use a preexisting client or server or implement their
> own , possibly one with a fancier interface , the possibilities are
> limitless.

Ironically, due to the aforementioned spam issue, something tells me that a
successful Usenet renaissance would yield one of two related solutions.

The first variant would be something like Mimecast or Mailprotector - users would
pay a company to implement spam filtering and 'good stuff prioritization'. This
has some advantages, in that services could compete on the efficacy of their
filtering solution, and also that users would have greater control over the
algorithm while being able to say "show me everything" in a verifiable way.

The second variant would be something like Gmail: "Usenet access, complete with
antispam and good stuff prioritization!" Which, Google Groups essentially is.
This sort of solution would end up being Twitter with extra steps. If Google were
to implement their Gmail filtering to their Usenet service, you're right next to
censorship.

The last variant is what you talk about below: having a myriad of servers users
can choose to subscribe to, and leave it up to the server ops to pick things to
remove. I'll address this below the section...

> You are a visionary so let me a suggest the following vision : every city
> block in every city in every technologically advanced country will have at
> least 1 usenet server operating. Note that the servers do not need any
> special facilities , it could just as well be a server operating from
> one's own home. It doesn't have to be a recent or powerful computer either
> , an old computer which one has lying somewhere and remains unused , would
> do the job just fine. The important thing is that all of these servers
> would be operated by different people. So lets say someone does not want a
> certain usenet group or messages on their server because they consider
> them as too right wing or too left wing or too whatever wing or they feel
> it's "hate speech" , etc. Not a problem. With so many servers the messages
> would still get transmitted between the people who are interested in them
> because there would be billions of different paths (passing through
> servers) between clients a message could follow. Now *that's* freedom of
> speech.

I don't think the lack of NNTP services is truly a problem:
https://www.reddit.com/r/usenet/wiki/providers/ For good or for ill, they don't
censor much of anything. As many of the existing Usenet services cater primarily
to binary downloads, the closest thing the existing companies seem to come to is
to handle DMCA takedowns. A handful of individual newsgroups are moderated, but
post removals on those aren't performed by server owners.

The sort of solution you're describing is either obscenely time consuming for
humans to perform (there are over 110,000 existing newsgroups), or those server
ops are stuck running a spam/content filter of their own and not letting end
users weigh in. Philosophically, is the solution to censorship "lots of different
censors to choose from"? Practically, can users on two federated servers have a
meaningful discourse if either one of them has a server op who deletes one half
of the conversation? We're back to the shadowbans of Twitter, but with two
potential chokepoints.

> So where do you come in ? No , I'm not suggesting that you pay for all
> those servers out of your own pocket. What you can do is express publicly
> your interest and support for usenet. Coming from someone as well known as
> you , this already will have a large positive impact. It may be that this
> is all that is needed. The infrastructure already exists : there is news
> client and news server software for the usual desktop operating systems ,
> there exist both free and commercial servers running and more can be
> created whether one is primarily motivated by profit or the desire to
> offer a public service.

I submit that there is plenty more that is needed. As you correctly point out,
all of this is in place already. This very discourse proves that. Tweaknews has
basically solved this. Their first party Usenetwire client is almost as simple to
use as Facebook (though some folks lament its terrible formatting), and €2 gets a
10GB block, i.e. more data than one could ever use for text-based discourse in a
lifetime. And yet, people aren't flocking to it.

Usenet has different sets of issues than Twitter. Creating a new newsgroup seems
more complicated than it needs to be, but paradoxically, there are many
newsgroups that are redundant and likely could be consolidated. Similarly, there
are swaths of abandoned newsgroups that haven't had a non-spam post since 2006,
and there's the awkward discussion about what to do with newsgroups that have
served their purpose (alt.windows95, anyone?).

The ability to avoid censorship by changing one's username and e-mail is
laudable, but it also means that genuinely bad behavior can't really be
regulated. Head over to alt.windows95 and look at the post from June 25,
2020...and let's try and figure out a solution for it. Deleting the post is
censorship, banning the author is basically impossible, and leaving it up there
validates the "why am I here" and "is this the upside to the absence of
censorship" questions that a whole lot of people would have. You and I can 'just
ignore it', but that's not a benefit to most people who would come by to look
around.

Twitter allows for pictures and GIFs to be part of posts, for good and for ill.
Usenet is still inconsistent with Unicode.

Usenet's asynchronous nature is helpful in that one needn't worry about missing a
post from last week. However, imagine the Twitter Users who already have a
tendency to mob and bandwagon getting infinite retention. It would make a flame
war last longer than it needed to, only for someone new to scroll up a bit and
restart the fire all over again. However, the paradox I find myself in as I write
this is the functional gatekeeping that the implicit alternative ("don't let them
on Usenet") recommends. I don't want to do that, but Usenet + Twitter Mob strikes
me as the worst of both worlds.

The presence of different Usenet clients has its benefits, but is also a
liability. Looking at Wikipedia's list of desktop Usenet clients, how many see a
meaningful amount of active development? Claws Mail, Thunderbird, and Seamonkey
were the only three I saw that had even one release in 2022, and none of them
have NNTP as their primary function any more than Chrome or Firefox were FTP
clients. Maybe one or two more applications on the list got updates in 2021, and
while I can appreciate Usenet clients avoiding a lot of the modern design cues
that software seems to focus upon, many that I've used seem to go too far in the
other direction. Of the thousands of desktop monitors I support in my job over
the past decade, there has been exactly one still running at 1024x768; the
overwhelming majority being 1920x1080 or higher. 16x16 toolbar icons are worth
revisiting. Usenet's connect/download/disconnect paradigm hails from the dial-up
or era; I appreciate UsenetWire's default download-on-select behavior. The
"Unified Inbox" in most modern email clients allows me to see recent mail from
all mailboxes in a single list. This should be a far more common option in NNTP
readers, but it's not. While SabNZBd and NZBGet and JDownloader get plenty of
active development reflective of their relatively high user count,
text-discussion clients seem to be subject to the other extreme.

Culturally, the handful of remaining Usenet post creators are of a particular
breed. We're generally tech savvy and generally can have a discussion that runs
its course and lets it sit. We can have a discussion over the course of days or
weeks, and it's fine. We can handle the technical issues and slower pace. Modern
social media and its users are unlikely to fit into such a culture, becoming a
bit of a chicken-and-egg problem.

> If you want to go further than that , you can announce a public competition
> for a new news client or a new news server with a prize for each of say
> 20,000 USD. That's not to say that there is anything wrong with already
> existing software but people get attracted to the new sexy thing and such
> new software , with the attendant publicity , would be the new sexy thing.
> I won't go into details of how such a competition could be run or what
> criteria should be used to rank the entries because it would make this too
> long and it would be a digression. The publicity would be more important
> than the precise rules of the competition anyway.

I addressed a lot of this already, but I submit that a grassroots return to
Usenet is going to be difficult to execute. Even if a few sexy newsreaders and
some additional servers were to be spun up, the differentiator you're proposing
over just being another Highwinds node is the huge number of different
moderators. That's what needs to be incentivized, which paradoxically, means
providing a financial incentive to censorship...which Twitter already has. I'm
all about giving Giganews some competition, but it's unclear how Usenet's
problems are solved by the presence of more servers, and even if moderation was
tied to servers, the problem has more to do with getting people dedicated to
performing the moderation on a continuing basis. Bandwidth and server maintenance
also play into the underlying question about how dedicated the grassroots sysops
would be. Some would be a 'labor of love' for a retired person, sure, but if they
get popular, it's a lot of work, and if they don't, it's work done in vain.

> You could create a charity which runs news servers. Again , the main
> advantage of such a thing would be the publicity caused by the association
> with your name rather than having a few extra servers.

Would the charities then be responsible for moderation the way grassroots servers
are? It's unclear how that's helpful.

> None of the above suggestions would cost more than thousands of USD. A lot
> cheaper than Twitter and they would do a lot more for freedom of speech. I
> will admit though that if your main goal in the endeavour is to make
> profit rather than enhance freedom of speech then a centralised medium
> offers more opportunity.

Although I agree with the statement expressly stated here, this goes all the way
back to Musk being able to accomplish 99% of what you're talking about with a
Mastodon fork, possibly with some volunteer moderators, and having a better
experience for everyone in the process.

> So these are my suggestions , I hope you will get to read them and give
> them some thought.
>
> Finally , for your convenience here is a list of some usenet servers :
> news.aioe.org , news.cyber23.de , news.eternal-september.org (requires
> registration) , news2.informatik.uni-stuttgart.de (read only).
>
> Best regards
> Spiros Bousbouras
>

I'm obviously not Elon Musk, and I do appreciate the fact that the problem is
being considered by someone besides me. I like the idea of more servers, more
grassroots servers, and even if there are occasional out-of-control flamewars, a
renaissance of Usenet would be worthwhile. As IRC got extensions and successors
in the form of Slack and Discord, so too could Usenet evolve in some way that
brings the best aspects of Usenet (client/server, organized discussion threads,
moderated/free-for-all choices, low bandwidth, asynchronous discussions with
splinters, etc.) while mitigating the worst (spam, total absence of even
light-touch moderation, no communities, no upvote/downvote system). I love the
handful of folks like you and me who still check in, but I think that bringing
the masses back here would ultimately be as much a fool's errand as buying
Twitter.