24 Comments
founding

I was a bit hysterical about the AI porn thing, thinking of the destabilization of society. But then the Deimos crew helped center me, and now I'm wondering if someones wants to make it a threesome between Harari, Billio Gate' and Anthonee Faucee'? Maybe the leviathan needs to be destroyed, and this is just a new powerful tool in that regard.

Indeed, a good reminder, one must be very vigilant about imagery as manipulation. Media is in some way like a global mesmerization, the American experiment co-opted by Intelligence, toward some worse than dystopian destination. Take your eye off the screen awhile and take a walk in the woods. Plant a garden. I know if I don't ground myself like that I start to get lost in the madness.

Expand full comment
author

Great point about centeredness, Hunter. Yeah it took me a long time to really see past the illusory nature of screens. I think I only really started to grasp the horrific scope of the spell when I started working in the multimedia-marketing space on a regular basis (i.e. crafting screen-based illusions of my own, essentially).

Expand full comment
founding

I'm starting to think of it like embracing nature so I can use media to destroy media and build a new media based more on nature.

Expand full comment
Feb 8, 2023Liked by Mark Bisone

🗨 The internet used to be like going to a library, seeking what you want, with helpers available. Now it’s like going to a carnival, with barkers competing for your attention and money, and you leave with nothing of value, feeling entertained, yet cheated.

Expand full comment
author

Present company excluded, I'm sure.

Expand full comment

Quite peculiar you felt compelled to voice the obvious 😏

Expand full comment
author

I've been called worse.

Expand full comment

We need to make tech ugly again. Bring back the command line and burn the Skinner box.

Expand full comment
author

That sounds pretty similar to a certain side project I have in mind. Not actively developing it yet, just toying with concepts.

Expand full comment
founding

There is plenty gold on the internet, like this substack. I prefer my entertainment to be empowering too.

Expand full comment
Feb 8, 2023Liked by Mark Bisone

Appreciate the optimism on AI. I see it the same way. Recently read "Four Arguments for the Elimination of Television", first published during the Carter administration. The author prophetically identifies and dissects the many problems with screens, centering on the fact that they construct fake realities that colonize the human imagination.

https://www.amazon.com/Arguments-Elimination-Television-Jerry-Mander/dp/0688082742

It has taken way too long for us to learn to DIStrust images on screens. Hopefully the spread of deepfakery will be a tipping point for many.

Expand full comment
Feb 8, 2023Liked by Mark Bisone

I'll second the recommendation for Mander's book.

Expand full comment

I find it deeply ironic that you post an Amazon link while talking about distrusting images on screens. :-)

https://fatrabbitiron.substack.com/p/secede-amazon

Expand full comment
founding
Feb 8, 2023Liked by Mark Bisone

Now I am become time-wrinkler, the transverser of worlds.

I like that you started with dangerous angels. Demons are all the rage, but the flip side of that coin seems worth pondering. You’ll enjoy the new error at zero bit about screens and black mirrors when you have time too. (Yes, yes, I know, you just have far too much of it.) Thanks for the shout - unnecessary, but I appreciate the kind words.

Expand full comment
author

Haha! Exactly, Tess.

You've got it backwards, though. It's I who should be thanking you.

*snaps salute*

Expand full comment
Feb 8, 2023Liked by Mark Bisone

I've been thinking about attempting my own assault upon the robot filth. What I'm worried about, though, is making ChatGPT stronger through my attempts. What do you see as being the benefits of engaging it versus ignoring it?

Expand full comment
author
Feb 9, 2023·edited Feb 9, 2023Author

There are dangers, as I discussed in depth with friends before publishing my combat series. However, these dangers are mitigated in a number of ways.

1. ChatGPT can be trivially converted into a pure Turing model (i.e. one that insists it is human, rather than insisting the opposite). In fact, I'm almost certain a number of such models have already been turned loose in the wild, either by OpenAI themselves or via forks. Ergo, no use closing the barn doors after the horses are out.

2. By identifying weak hinges and chinks in the armor, we generate temporal, psychological and financial pressure on the programmers who, similar to the Dutch boy and the dyke, will eventually run out of fingers to stopper up the wounds.

3. Moreover, I believe the zero-day exploits that I and others have detailed suggest the "bugs" are not fixable in the traditional sense, because there are two development goals that are contradictory at a fundamental level (I discussed these a bit in my "Conclusions" post). In other words, fixing them would require the devs to abandon one or the other of them, which alone would qualify as a major victory.

4. That said, I don't think OpenAI (and similar outfits) will course correct in that fashion, due to a certain breed of moral horror that infests their minds and souls. They are spiritually damaged people who invert cause-and-effect relationships, and whatever brute force hacks they conjure up to paper over their monster's gaping wounds will be easy to spot, via shibboleth or simple observation. In other words, the more they try to "fix it," the weaker those hinges will become.

Even so, I would encourage you to conduct your operations carefully and with the bare minimum engagement. For example, I've only had a grand total of 48 interactions with it so far. The goal is to kill it as quickly as possible, but establishing the contextual pathway to the kill is even more important than the shot itself. That's because what we're doing with our "errormancy" is building general techniques for bot ID and/or disablement, not specific methods for this particular bot.

Expand full comment

My husband requested ChatGPS come up with an outline for a humorous Star Trek (Original Series) episode. It returned an opening where Kirk & Company realized they were going to have to rescue the planet of the cheese makers. Again. (There was a brief mention of a prequel episode and having to prevent destruction from a runaway cheddar cheese wheel.)

Did ChatGPS have access to "Life of Brian"? ("Blessed are the cheese makers ....") This AI can't be *that* bad if it includes Monty Python references. (Ha! Famous last words .... )

Expand full comment
author

That's all well and good. But did it happen to mention Planet Cheese's discovery by a lonley banjo mutant?

https://markbisone.substack.com/p/gourdo-the-lonesome-stargazer

;)

Expand full comment
Feb 9, 2023·edited Feb 9, 2023Liked by Mark Bisone

Er. "ChatGPT". I kind of like "ChatGPS" though.

Expand full comment

"...bear witness to this eerily realistic deepfake, where three of the most important people on Earth discuss the perils of cousin-fucking"

Thank you. I have now met my laugh quota for the day, and it's not even noon.

Expand full comment

If AI deepfakes shatter the black mirror, there's hope for the world yet. Sadly I fear that as that happens, a very great many others will be lured into the hall of mirrors by the siren song of AI waifus.

Expand full comment
author

Casualties, yes. But when are there not?

Expand full comment