iiinova — where your digital self keeps evolving

Theodore Morley @morleytheodore.bsky.social · 28/09/2026

These are certainly very reasonable worries. Thanks!

Theodore Morley @morleytheodore.bsky.social · 28/09/2026

Makes sense, yeah. I think ambient levels have definitely increased, but also that they've increased unevenly depending on what spaces people inhabit online. What form would you see that event taking?

110

Theodore Morley @morleytheodore.bsky.social · 28/09/2026

I think on some level we're already there re: trust of information, the trustworthiness of the internet is already gone compared to what it was at one point, largely due to the ease and scale of manufacturing natural language discussions, images, and other content to support chosen viewpoints.

100

Theodore Morley @morleytheodore.bsky.social · 28/09/2026

A significant shift in my mind was the mid-2010s spread of engagement-based recommendation engines (rather than from follows). It makes things much more engaging, and also removes any real function for tastemakers or reviewers, meaning the gate to engage with media at all is completely non-existent.

000

Theodore Morley @morleytheodore.bsky.social · 25/09/2026

This feels like an incredibly alien take. Why would anyone care about this at all?

010

Theodore Morley @morleytheodore.bsky.social · 24/09/2026

Thinking back to what I've read about early 1900s debates around vehicles, stricter liability for drivers after they injured people in the course of driving would probably have had long term downstream affects on city planning, I'd guess.

010

Theodore Morley @morleytheodore.bsky.social · 21/09/2026

Personally, maybe this is a simple one, but in medical fields the unstructured notes data has always been very information rich but also infeasible to access information from at scale, but cheap and reliable processing for it potentially opens up a lot of opportunities for research programs.

130

Theodore Morley @morleytheodore.bsky.social · 21/09/2026

I really didn't expect philosophy of mind to be the college class I took which would be most relevant to public discourse in a decade.

000

Theodore Morley @morleytheodore.bsky.social · 09/09/2026

I think the most parsimonious explanation is that a lot of people don't really know what alignment is and just read the statement as "as a higher-up at big AI company"

041

Theodore Morley @morleytheodore.bsky.social · 09/09/2026

Also, this is something that I assume OpenAI would be able to verify, just the question of whether a given account has opted into training or not. My main confusion is that this seems to not be stated anywhere by any parties in this, though maybe I missed something.

100

Theodore Morley @morleytheodore.bsky.social · 09/09/2026

I suppose because there isn't any public artifact I can find describing what their deid process is, I'm not entirely sure what's happening there. But I would be curious to know whether the "use to improve our models" toggle was on for his account or not.

100

Theodore Morley @morleytheodore.bsky.social · 09/09/2026

This is model to model dependency graphs, which makes sense, but the artifact in question should be within their own corpus, with association to any internal model training regimes or inputs. They have contractual obligations on enterprise data to verify this. But maybe deid breaks the chain here.

110

Theodore Morley @morleytheodore.bsky.social · 09/09/2026

Separating two claims: Figuring out if a given piece of training data influenced output is obviously hard. Checking what was in the training set should not be hard. Apologies if unclear.

120

Theodore Morley @morleytheodore.bsky.social · 09/09/2026

I'm curious what you would find shady? I haven't felt that. I will say I find it somewhat alarming that OpenAI is claiming that it's infeasible to do a simple set membership check on training data inclusion. That should be fairly well tracked, shouldn't it?

120

Theodore Morley @morleytheodore.bsky.social · 09/09/2026

That's fair for context. I feel as though the core issue here is the seemingly acrimonious relationship between labs

110

Theodore Morley @morleytheodore.bsky.social · 09/09/2026

He makes that claim in the section where he's talking about what he says he wanted to propose though, not anything he actually said. The "it would be simpler" part is referring to the prior proof, right? I feel like they could work jointly with acknowledgement of which models were used where.

110

Theodore Morley @morleytheodore.bsky.social · 09/09/2026

I guess so, but I'm still unsure about what is going on there. Even the second point still seems to be a veiled implication that the writer didn't want an Anthropic employee as a coauthor. Doesn't that somewhat admit that the business interests were driving that statement?

120

Theodore Morley @morleytheodore.bsky.social · 09/09/2026

Does he address the claimed "If you don't want me to be nice, then I don't have to be nice" quote? Seems out of context with his claim about the context of the "ruin your career" bit.

121

Theodore Morley @morleytheodore.bsky.social · 31/08/2026

- via varying social media formats, then I would expect the share of that read on the overall economy to, over time, start to have larger parts of it determined by what people who you don't know say, potentially a non-representative set algorithmically surfaced based on engagement metrics.

030

Theodore Morley @morleytheodore.bsky.social · 31/08/2026

It would honestly be surprising to me if it wasn't. Just if we think in the abstract, if an individual's read on the economy is informed by: personal situation + news + situation as described by friends, and then you add in a bunch of non-friend people in the same source of "described by friends"-

150

Theodore Morley @morleytheodore.bsky.social · 25/08/2026

Terrible and heartbreaking news. She lived life wonderfully, few people can say they created something as incredible as the Imagination Library, or any other of the many amazing things she's done.

000

Theodore Morley @morleytheodore.bsky.social · 22/08/2026

Just the other day I was saying this intersection I drive through regularly really should be an all way stop instead of just on the one side because of the hill being dangerous. Then I check and it is! Pretty great to see the street get safer. Thanks @freddieoconnell.bsky.social !

010

Theodore Morley @morleytheodore.bsky.social · 17/08/2026

Amazing book, I loved reading the Wager. Definitely my favorite book that I read in 2024. One favorite scurvy fact was how some of the only people on ships who didn't suffer from scurvy were the men locked away eating rats, due to the rats' capability to synthesize their own vitamin C.

121

Theodore Morley @morleytheodore.bsky.social · 28/07/2026

Another element is that the character does look rather like a person walking, so it made for an effective marketing campaign by Yamasa.

0100

Theodore Morley @morleytheodore.bsky.social · 16/07/2026

Like, obviously a disease that will be captured as a positive test when symptoms present basically anywhere in the country has very different measurement dynamics than a disease for which standard diagnostic testing will return negative results unless you order a specific workup for cyclospora.

020

Theodore Morley @morleytheodore.bsky.social · 16/07/2026

This piece is really bizarre. She clearly sees the evidence that there isn't adequate testing, that we don't have accurate case counts, and then proceeds to use that data to suggest reassurance. Comparing endemic diseases that are on standard stool panels to outbreaks that aren't captured is odd.

120

Theodore Morley @morleytheodore.bsky.social · 16/07/2026

When I was in high school, one of the teachers casually told me about how the area with couches where students could hang out used to be the smoking lounge for students back in the day.

030

Theodore Morley @morleytheodore.bsky.social · 14/07/2026

If they get no punishment for murder, but the people protesting the murders get punished simply for pointing out the violence, it should actually be covered as an event which is much, much worse than a serial killer.

000

Reposted by Theodore Morley

Micah @rincewind.run · 14/07/2026

this is the closest ICE has ever come to saying "yeah we just murdered a guy"

283151690

Reposted by Theodore Morley

Max Kennerly @maxkennerly.bsky.social · 14/07/2026

The best story ICE could make up, after 12 hours, is that ICE agents shot someone in the head because they believed the person they shot, who was not accused of any crime, was a different person also not accused of any crime, and so "public safety" required killing them on the spot.

8840691454

Theodore Morley @morleytheodore.bsky.social · 02/07/2026

It's a little known fact that all US presidents can access the avatar state

110

Theodore Morley @morleytheodore.bsky.social · 16/06/2026

In addition to the story-games and the OSR groups, I remember all the PBTA type crew was very much centered around g+ in the like mid 2010s time period.

040

Theodore Morley @morleytheodore.bsky.social · 16/06/2026

Yes, Google+ was actually extremely popular in the particular subgroup of TTRPG designers at the time, it had some surprisingly good features for TTRPG discussion. You could of course embed Google Drive docs natively into posts, which was of course very useful for people talking about rulebooks.

040

Theodore Morley @morleytheodore.bsky.social · 04/06/2026

If interested in this, one of the primary works is "Fighting Traffic: The Dawn of the Motor Age in the American City" by Peter Norton.

000

Theodore Morley @morleytheodore.bsky.social · 04/06/2026

That's honestly a pretty accurate descriptor of what the early automobiles were like. There were very significant public arguments about the fact that the (very wealthy) people driving automobiles in cities would often run over children, pets, and the elderly. The car industry invented "Jaywalking".

100

Theodore Morley @morleytheodore.bsky.social · 21/05/2026

Hey what did Nashville do to get brought into this whole beef

insult deflection meme about the user insulting nashville as an attack on weird billionaires but instead hitting normal people who live in nashville

010

Reposted by Theodore Morley

Ari Berman @ariberman.bsky.social · 07/05/2026

Breaking: Tennessee Republicans pass new 9-0 gerrymander after SCOTUS destruction of Voting Rights Act eliminating state’s only majority-Black district & last Democratic seat. Splits Memphis, which is 63% Black, into 3 districts to disenfranchise Black voters. Quite literally a return to Jim Crow

20073043069

Reposted by Theodore Morley

The Tennessee Holler @thetnholler.bsky.social · 07/05/2026

⚡️ WATCH — JUST NOW — 99% white Tennessee House Republicans pass a racist 9-0 map stripping majority Black Memphis of congressional representation

2441604946

Theodore Morley @morleytheodore.bsky.social · 07/04/2026

This person should not hold any government office. This, along with innumerable other things, are clear evidence to any reasonable observer of his insanity. The other members of the United States government hold the power to remove him from office and have not. This is a clear moral failure.

000

Theodore Morley @morleytheodore.bsky.social · 18/03/2026

I remember being fairly horrified at the time reading about the permanent brain damage he suffered afterwards.

020

Theodore Morley @morleytheodore.bsky.social · 15/03/2026

This is one of the hardest parts about working in health research in general. So much of the time, the clearly large changes that can be made to improve outcomes are already completely known, we simply don't do them for political reasons. No comment on the dog vaccine, haven't read enough on it.

011

Theodore Morley @morleytheodore.bsky.social · 12/03/2026

Why would a convoy of tens of thousands of trucks driving in a straight line be *less* vulnerable to disruption than the oil tankers? How would you even get that many vehicles assembled in the area? I feel like people just refuse to think about how scale affects things.

020

Theodore Morley @morleytheodore.bsky.social · 12/03/2026

These places are all connected by land already to this person's proposed pickup point! There's already pipelines doing this in essence! If it was viable to move that oil by land, you don't think they might have tried that in the last century? You would need 10,000 truck drivers for each tanker!

110

Theodore Morley @morleytheodore.bsky.social · 12/03/2026

Listen, there's a lot of reasons why this is an incredibly stupid idea. The logistics of moving load from a tanker (2 million barrel capacity) to trucks (200 barrel capacity), the number of people needed, drones, pickup point being a village... But one easy visual is to zoom out the map a little.

The region surrounding the Strait of Hormuz, showing that the area they're talking about having as a dropoff point contextualized by the surrounding area and shape of the Persian Gulf/amount of land controlled by Iran.

140

Theodore Morley @morleytheodore.bsky.social · 10/03/2026

I didn't really think this blog post I wrote last year was gonna be relevant again so quickly tjmorley.com/blogposts/co...

tjmorley.com

Cottagecore Programmers: The Idealization of Farming by Tech

010

Theodore Morley @morleytheodore.bsky.social · 10/03/2026

But this sort of labor is truly physically demanding in a such a way that casting it as some sort of self-actualization experience, divorced from the actual reality of that labor, is somewhat bizarre.

0132