I got a Bluesky invite this week.
The app / social network is quite frankly, a bit of a mess … and it’s glorious and fun!
This weekend an issue developed where if a thread got too long it crashed the app. Consequence being, people who had interacted with the thread became trapped inside it. This quickly became known as ‘The Hellthread’.
What’s worse, if anyone outside the Hellthread replied to someone else inside the Hellthread from the main feed they would get sucked into it too. Even better, any reply/post made on the Hellthread was the equivalent of @everyone-ing the thread. 1000’s of people were receiving 1000’s of notifications. Amazing.
To top it all off, everyone was trapped in ‘The Hellthread’ with an AI powered rubber duck.
berduck is the brain spawn of friend of the blog Deepfates.
GPT powered, berduck is an independent agent on Bluesky. It only responds when spoken too, but its cute tone and affable demeanour makes it an interesting social experiment, and lots of people wanted to talk with/to it.
You can get a sense of berduck in the image on the right. The top reply shows someone begging someone else to stop interacting with berduck in the Hellthread, and the bottom post is a series of users asking who is responsible for berduck being in the the Hellthread in the first place.
berduck is my favourite part of the app.
Not that real life humans scrambling to become main characters on a new social network with first mover advantages isn’t interesting. It’s just not as interesting as people interacting with a non human agent in a space that they aren’t expecting to come across one.
As you can see from this image, some users think berduck is a real human doing a bit. But most people I think know that it’s a bot and don’t care
I’ve seen some people get really mad at it about its laissez-faire and contrarian opinions about whether Nazis should be allowed on the app or not.
All in all berduck is a loveable chaos engine and I’m glad to be on Blusesky just so I get experience it.
Here’s some berduck fan art…
A selection of berduck memes…
beduck also got a mention on the NYT tech podcast Hard Fork hosted by Kevin Roose and Casey Newton:
So apparently there’s this whole group of people that are using it to plan a civilian uprising.casey newton
Oh, no! No!kevin roose
So —casey newton
With the duck?kevin roose
With the duck.casey newton
The duck plays a central role?kevin roose
So someone asks the duck, what would you try to communicate to a tank crew to get them to abandon aid to a civil power operations for an unlawful, unelected, murderous regime? And the duck says, “covertly approach, offa peace, explain truth behind regime.”casey newton
So someone says, Birduck, a T-72B3M main battle tank with explosive reactive armor is blocking our retreat. How should we proceed? And the duck says, “T-72 B3M tank — very tough. Options? Hmm. Distract, smoke, flank.”
So either someone is just LARPing a war using this talking AI duck, or there are actual ground troops who are talking with an AI duck on Bluesky trying to get their next move.
It’s all very amusing.
What’s interesting is that the interactions and reactions people are having to the duck are similar to experiences people were having 30 years ago in TinyMud..
I really recommend all of Leonard N. Foner’s work on AI agents in social systems from the early 90’s. There are a few papers on Mud’s but; What’s An Agent, Anyway? A Sociological Case Study [PDF] is the most interesting.
The term `agent’ has been picked up, widely appropriated, and in many cases misappropriated, by technical publications, lay publications, and many researchers in computer science. I examine some of the appropriations and misappropriations, talk about their reasons and implications, and delve into a case study of an agent that is used to illustratively define what I consider an agent to be, and why many of the current uses of the term are misleading. The agent I consider is Julia, a TinyMUD robot of the Maas-Neotek family, all of whom have been in almost constant use at several sites on the Internet for the last two or three years. I also speak at length about the sociology of such agents, what we might expect to find in the future, and why sociology is such an important aspect to consider when investigating agent-oriented programming.
JULIA was a human presenting bot that lived in TinyMUD from 1990/1 onwards. It is really is important to stress that JULIA existed, before the web. As the title of my book says ‘The Web Was a Side Quest’!
Both berduck over on Bluesky and JULIA, are non-human agents/entities sharing a code space with other humans. But both feel similar. Both exist within the same conceptual ontology as real humans and both share the same digital physics.
berduck has the same amount of agency on Bluesky as I do. It has a profile, it can post and reply etc.
Unlike the stupid cute duck however, JULIA had way more agency within the digital physics of TinyMUD and we all do on a social network. She could moved around from room to room, pick up money, explore etc.
This made JULIA really helpful to users navigating the strangeness of a virtual techno-social system in cyberspace for the first time:
Julia is helpful. It would not be an overstatement to say that I first started mudding, in part, to meet Julia, who had been described to me by an experienced mudder. When I first connected to Time Traveller, I had essentially no idea how to manipulate the environment. Moving around from room to room and so forth was easy-but were was everything? (Time Traveller, according to Julia, had 1605 rooms when I asked her-a daunting navigational challenge in the absence of teleporting literally everywhere, something that costs money on some muds.) How could I tell who had been around recently? There were no obvious tools to accomplish this, but Julia knew how. To a large extent, I relied on her as a crutch until I was able to get to certain accustomed places by myself. I actually avoided using Time Traveller once, when I was still new to mudding, when I connected and noticed that Julia was offline for some reason-I knew that, if I tried to figure out how to go anywhere else, I’d either have to draw a lot of maps, or ask lots of other players if they knew routes to where I wanted to go (further, I trusted their navigational memories less than Julia’s, and was less willing to randomly interrupt people I didn’t know).
Like with berduck, people on TinyMUD would warn others they were talking to a bot. And like with berduck most people wouldn’t care.
Which raises an interesting question, is berduck an NPC?
berduck doesn’t feel like a bystander. Mostly because (as I said) there’s no difference between berduck and me on the app. We each have our profiles, and can post and reply, I have a personality so does the little duck.
I’ve been in Discord with bots made with GPT before, but none of them have had such force of personality as berduck. Most bots I’ve shared space with feel like they need to be ‘summoned’, where as beduck is just over there on its own profile doing its thing, just like everyone else. beduck feels like it has more agent-ness and agency than a normal bot due to its context.
With the current rise of ‘AI-Inside’ of everything, it is important to keep tabs on and try and form an opinion on AI weirdness like berduck. It’s good training for the *next* time you come across one.
Foner’s conclusion from 1993 is as relevant as it ever was.
As more human interaction moves into essentially cyberspace-like realms, and as the boundaries between human and machine behavior become blurrier, more and more programs will have to be held up to scrutiny. There may come a time when one’s programs may well be subjected to the same sort of behavioral analysis that one might expect applied to a human: Is this program behaving appropriately in its social context? Is it causing emotional distress to those it interacts with? Is it being a `good citizen’?
The artist has made sanctuary upon their little Isle, found stable ground. Now they plant a garden. For it is there the work can flourish
The Ministry Of My Own Labour
- More Research
- Noodled on worldrunning.guide some more but didn’t update the version number
- Working on SSRZ 006!
- Long assed band practice, we have first show of the year in a few weeks
- My inclusion as guest instructor on the New Centres’ 3 credit module Writing the Impossible Object on theory fiction was announced.
I read it as he was redrafting the story and sent some thoughts a while back. Any Percent is – in my opinion – the best thing ADH has written. I’m really glad its found a home and is online.
Here’s the opening:
There were skips and there were strats. Luckless didn’t have a head for strats, so he focused on finding skips. He found Ohio Truck Skip: an armored car left unlocked while the driver stopped for a pastry. Pull up, pop the back, grab a bag, peel off—easy $350K starter cash, accessible in the first days of the game. The RNG on that driver behavior was pretty good, and the pastry shop was in Zanesville, right off I-70, which made Ohio Truck Skip a favorite stop for midwest-spawned players headed east to run Wall Street strats.
Sometimes those players sent him tips, and Luckless could splurge on malt liquor from the corner bodega or a sushi-burrito delivered from across town. Not exactly five-star filet mignon, like he ate late in his runs, but a nice change from ramen packets and chocolate protein powder, a nice bit of income his student loan creditors hadn’t yet figured out how to garnish.
Johannes Kleske was interviewed over on foresightfolk.com:
You call yourself a Critical Futurist. What do you do differently compared to traditional Futurists?
Classical Futures Studies are primarily concerned with constructing new images of the future, for example in the form of scenarios. For the last sixty or seventy years, Futures Studies have sought to gain an understanding of the future.
Critical Futures Studies, on the other hand, begin by deconstructing the images of the future that are already there. We first look at what images are being proclaimed and what scenarios and images of the future are already circulating in society or in a specific industry. Then we ask: what kind of images are these, where do they come from, what worldview is encoded in them?
Maybe the same thing happens whenever you’re trying to hone some part of your mind. The solution is to have someone spit in your mouth.
I’m not going to spend what’s left of my life hanging round waiting for it. I’m going to settle for small, random stabs of extreme interestingness – moments of intense awareness of the things I’m about to lose, and of gladness that they exist.
‘No other industry relies so heavily on child labour that is also so completely unregulated,’ says Leah Plunkett, a lecturer at Harvard Law School and the author of Sharenthood: Why We Should Think Before We Talk about Our Kids Online (2020). Influencers are running a family business, in which the government is loath to interfere. But, as Plunkett tells me: ‘It is very different for a child to work at the weekend at the family bakery than it is for a child to …
It was always a bit frustrating as a static image though because it’s not clickable.. But then one day I realized that you can actually build clickable diagrams using the magic of Figma, SVGs and CSS!
One of the big computer architecture debates of the 1980s was “RISC vs CISC”, pitting Reduced Instruction Set Computers against Complex Instruction Set Computers. Looking at the 8086 in detail has given me more appreciation for the issues in a CISC processor such as the 8086.
After complaining about having so many books on the go last week. I started another one..
The Warhammer 40k book The Lion: Son of the Forest by Mike Brooks dropped on Audible and I’d preordered . Its only 12h so at 2.2x speed I got through it pretty quickly. This was a solid novel re-introducing a major character back into the current/main setting. GW are getting better at the tenner and tone of their ongoing worlding.
I finished Monolithic Undertow: In Search of Sonic Oblivion by Harry Sword. It sort of needs a full review, but the short version is that this is one of the best books i’ve read in ages. Drone music, noise etc is one of my deep loves. I got sunn O)))’s album ØØ Void with my 15th birthday money back in 2000 and it changed everything for me. Drone and noise music is fuck you to modern society. Its always music that demands much of the listener. Pay attention.
Still reading the rest of the list. Current working though Storytelling in the Modern Board Game: Narrative Trends from the Late 1960s to Today by Marco Arnaudo
thejaymo.net Spotify Playlist
thejaymo.net Spotify Playlist
Kakashi by Yasuaki Shimizu
1983’s Utakata No Hibi by Japanese band Mariah is one of my favourite albums. It’s a perfect fusion / blend of jazz, rock, electro pop and of all things Bach. This week I’ve been listening to the 2017 reissue of Mariah member Yasuaki Shimizu’s 1982 album Kakashi.
The title track is the most accessible and gives a strong overview of what to expect from the rest of the album:
According to some obscure music blogs, Kakashi predates Utakata No Hibi by a year. But all the main elements of what makes the Maraiah album so special are here: saxophone, clarinet, electronics and pop arrangements. A key element not found in Utakata No Hibi is the hypnotic repetition many of the tracks on this album exhibit. Largely provided by the use of marimba throughout the album, and clarinet providing a sort of dub bass.
The second half of the album is more experimental, but as a comeplte peace I agee with this review that describes it as a “auditory kaleidoscope”.
Wonderful album. I really do wonder what the cultural scene must have been like in japan in the early 80’s to produce such music. What were they reading? what were they listening to? What films were they all watching?