iiinova — where your digital self keeps evolving

Kevin Riggle @kevinr.free-dissociation.com · 6h

Fifth Law of Procrastination: Procrastination avoids boredom; one never has the feeling that there is nothing important to do. —fortune(6)

081

Kevin Riggle @kevinr.free-dissociation.com · 08/10/2026

Laura's Law: No child throws up in the bathroom. —fortune(6)

010

Kevin Riggle @kevinr.free-dissociation.com · 08/10/2026

Shouldn't that be an argument for increasing our standards rather than relaxing them?

120

Kevin Riggle @kevinr.free-dissociation.com · 08/10/2026

There's a lot of overlap with cybersecurity research too. "I wonder what happens if I push... *this*??"

050

Reposted by Kevin Riggle

Gert Sønderby @gertthenerd.bsky.social · 08/10/2026

A good QA has the imagination of a child and the deviousness of a spy, combined with lots of experience of how things break. They are invaluable.

1532

Kevin Riggle @kevinr.free-dissociation.com · 08/10/2026

I mean verified by humans. The production of knowledge is a human, social process. These haven't undergone that. (How many bugs in the Lean compiler are these going to turn up, I wonder)

140

Reposted by Kevin Riggle

Jacky Alciné @jacky.blacksky.app · 08/10/2026

Do I know anyone on here who works at a Ruby shop (Rails preferred but not strictly that) and needs some talented folks to help with performance? Asking for a friend (actually not me).

022

Kevin Riggle @kevinr.free-dissociation.com · 08/10/2026

I mean or they can go get their work peer-reviewed the way we normally do science. Until they do, it's exactly what it looks like—unverified results

150

Reposted by Kevin Riggle

absolute horses @jcoglan.com · 08/10/2026

it's not that they themselves lack domain knowledge, they clearly understand the system in question in great depth, but the LLM is constantly invoked as an excuse not to do the work of conveying any of this knowledge to anyone else

1231

Kevin Riggle @kevinr.free-dissociation.com · 08/10/2026

It's the Curse of LISP all over again

010

Kevin Riggle @kevinr.free-dissociation.com · 08/10/2026

Fundamentally here OpenAI have the same epistemic status as the weirdos at the back of the physics conference with a fancy poster trying to tell anyone who will listen about their perpetual motion machine. OpenAI are just bigger

2196

Reposted by Kevin Riggle

jae (hallowed) @fubarchitect.com · 08/10/2026

this is absolutely a correct response. this is expert level gatekeeping imnsho. "i don't care, go away" is exactly how you respond to a million line pull request that fixes a bug your users have not encountered, which is what this is

4547

Reposted by Kevin Riggle

Jesse Felder @jessefelder.com · 08/10/2026

‘If AI doesn’t deliver on its promise then we face a double blow to growth: because AI didn’t expand our productive capacity as hoped, and because other investments which might have didn’t take place.’ www.ft.com/content/bf40...

48939265

Reposted by Kevin Riggle

Bhaskar Mitra | ভাস্কর মিত্র @bmitra.bsky.social · 08/10/2026

"Non-use is the default because we want use to be an intentional decision, reached through a reflexive process." This is a good policy document. More institutions / departments need to articulate explicit guidelines like this.

0204

Reposted by Kevin Riggle

Dr Abeba Birhane @abeba.blacksky.app · 08/10/2026

we have a policy for the use of genAI at the @aial.ie. we reflect on values key to scientific integrity, such as replicability, and err on the side of not using it blogpost: aial.ie/blog/gen-ai-... pdf: zenodo.org/records/2322...

A flowchart that guides reflective thinking around generative AI in the context of research. It highlights key, interconnected values underpinning scientific integrity, alongside practical and wider socio-political considerations when deciding whether to use generative AI for research.

520994

Reposted by Kevin Riggle

Charles Louis Richter @richterscale.bsky.social · 08/10/2026

witness the snoozing power of these FULLY VACCINATED KITTENS

Two kittens snoozing stretched out all over each other

Same two kittens snoozing in a slightly different configuration of cuddles

622527

Reposted by Kevin Riggle

Molly Gee @voolala.bsky.social · 07/10/2026

Day seven: panic

An ink and colored pencil drawing of a disco ball with a panicked cartoon expression. He has silver teeth and there is a poor approximation of colored disco lights in the corner. The words on the page read "Inktober 2026 October 7" and "panic" (the prompt)

3111

Reposted by Kevin Riggle

Erin Kissane @kissane.myatproto.social · 08/10/2026

It says we owe the dead more than we think we do but bloom in contact with the living world, and with other humans most of all. It says we are all beloved children of the House, which is how the unresolvable ambiguity of human experience is made beautiful and (therefore) survivable.

3816

Reposted by Kevin Riggle

Erin Kissane @kissane.myatproto.social · 08/10/2026

I was waiting for someone to pop in and say “I thought it was about the internet” so I could say “That’s what I said” but alas

2593

Reposted by Kevin Riggle

Erin Kissane @kissane.myatproto.social · 08/10/2026

Piranesi says the world and the epistemic bubbles we actually live in are prisons and but also liberation, that the terrible limits on cognition imposed by [religion, science, devastating illness] are also protective freedoms, that amnesia is salvific but makes us brutally vulnerable to abuse.

415928

Reposted by Kevin Riggle

Douglas Reyes-CeroÑ @dreyesceron.bsky.social · 07/10/2026

The Anaheim Ducks’ incredible organist Lindsay Imber played a rendition of A Cruel Angel’s Thesis from Evangelion during the Ducks’ home opener vs. the Florida Panthers www.instagram.com/reel/DeH0AXO...

2464211

Kevin Riggle @kevinr.free-dissociation.com · 08/10/2026

Yes and no, like everything. In some places it succeeded too well and now the thermostat needs to turn off again. In others its aesthetics got voopeted by management and we could take a heat gun to them

160

Kevin Riggle @kevinr.free-dissociation.com · 08/10/2026

ArXiv has accomplished a lot and will yet accomplish more but this is a dynamic optimization problem not a binary and we’re too far on the “flood the zone with shit” side

010

Kevin Riggle @kevinr.free-dissociation.com · 08/10/2026

People who don’t know what ArXiv or the Poincaré conjecture or the Riemann hypothesis or whatever is

200

Kevin Riggle @kevinr.free-dissociation.com · 08/10/2026

The whole point of agile is to take power away from c-suite groupthink’s and invest it in people closer to the customer

150

Reposted by Kevin Riggle

ve, the ghost of info past @vortexegg.com · 08/10/2026

Oh yes, ain’t that the truth. And specifically, I think it is about balancing not just the internal power of decision-making, but the internal power of theory building and collective sense-making; which is why C-Suite counter-knowledge groupthinks are the worst thing you can let happen in an org.

261

Reposted by Kevin Riggle

austin @aparker.io · 08/10/2026

oh 100% the funny sidebar to this is that waterfall/agile/devops/whatever are more about the balance of internal power between ICs and management than anything else

282

Kevin Riggle @kevinr.free-dissociation.com · 08/10/2026

Oh god yes. Microservices architectures are there to move middle management disagreements out of the ABI level and into the API level basically

130

Reposted by Kevin Riggle

ve, the ghost of info past @vortexegg.com · 08/10/2026

Interestingly this is where I came away with the understanding that microservices architectures are an organizational design pattern, not a software architecture one.

5191

Reposted by Kevin Riggle

ve, the ghost of info past @vortexegg.com · 08/10/2026

I’ve worked at an org that did that in the pre-LLM days. Most of the work I did for them was helping them effectively undo that decision (the issue was that all the internally competing teams and systems had to work together to be one company offering one single working product).

1141

Reposted by Kevin Riggle

ve, the ghost of info past @vortexegg.com · 08/10/2026

Yes, it’s absolutely a requisite variety problem, a la Beer’s reformulation of Ashby’s law in organizational settings. The fundamental cybernetic problem with LLMs is that they are *variety monsters*, which necessitate more and more technical management to be layered on top of them.

291

Kevin Riggle @kevinr.free-dissociation.com · 08/10/2026

I’m not saying you’re wrong, I’m saying I need more information

010

Kevin Riggle @kevinr.free-dissociation.com · 08/10/2026

*looks at Waymos* *looks at the current state of autonomous semis that turn into autonomous mecha*

110

Reposted by Kevin Riggle

Sharon 🪳🌹 @sharonk.bsky.social · 08/10/2026

So a Sharon-splainer of what's happening in India: Basically, in India, election rolls are a lot more centralized than in the US. As a result, the national electoral commission has been engaging in heavy voter roll reviews (primarily in opposition-friendly states) called SIRS.

223539

Kevin Riggle @kevinr.free-dissociation.com · 08/10/2026

My longstanding joke is that should any servers I’m responsible for develop sentience problems I will call up a backhoe operator and point them at the datacenter uplink

030

Kevin Riggle @kevinr.free-dissociation.com · 08/10/2026

They make a lot of us who are inspired by her and trying to carry her work forward though

000

Kevin Riggle @kevinr.free-dissociation.com · 08/10/2026

Also I think the highest and best use of genAI may be generating content for a free bar trivia event

050

Kevin Riggle @kevinr.free-dissociation.com · 08/10/2026

The guy who runs the bar trivia I go to had a kind of cute round tonight where you had to identify the two characters in an AI generated mashup e.g. (not a real example, don’t yell at him) Dirty Harry Potter It was, I thought, actually a decent use of genAI

150

Reposted by Kevin Riggle

ve, the ghost of info past @vortexegg.com · 08/10/2026

You can pretend that the agents told you they proved it, and maybe it actually coincided with reality, but that is an accidental hallucination on the part of the development team and not an intentional outcome.

1372

Reposted by Kevin Riggle

ve, the ghost of info past @vortexegg.com · 08/10/2026

More plainly, given a set of people who need to collectively understand something in a finite window of time (the pre-requisite for successful software), increasing the number of people inversely scales the time required, and using agents exponentially scales how much is required to make sense of.

27010

Reposted by Kevin Riggle

ve, the ghost of info past @vortexegg.com · 08/10/2026

Ultimately the problem is that you can’t scale up the social activity of theory building by either adding more people to the set of people who need to understand the theory, or automating the increasing surface of theory to be understood because collective knowledge has inverse scaling frictions.

311416

Reposted by Kevin Riggle

ve, the ghost of info past @vortexegg.com · 08/10/2026

We have invented the mythical man month from classic software engineering book Don’t Invent the Mythical Man Month.

3461109

Kevin Riggle @kevinr.free-dissociation.com · 08/10/2026

We’ve gotten better but also our reach has so far outstripped our grasp

000

Kevin Riggle @kevinr.free-dissociation.com · 08/10/2026

Spamming datacenters across the landscape

010

Reposted by Kevin Riggle

Luke Dicken @luked.bsky.social · 08/10/2026

This is the tweet. On every axis, GenAI is a DoS attack against society. "We can generate more of everything than you can possibly parse" - clogging up every system. Destroying every social contract. Up to and including discourse and even fucking construction.

1184

Kevin Riggle @kevinr.free-dissociation.com · 08/10/2026

You share this by publishing it in reputable peer reviewed journals

110