iiinova — where your digital self keeps evolving

Reposted by Joe Simons

Dr Farbod @emergencybod.medsky.social · 28/09/2026

And yes, part of me wants that moment to be a gigantic, continental-sized "fuck you" to everyone who told my fellow Britons to “go home”. We are home. And I want my country back too.

2536

Reposted by Joe Simons

Beau @drbeaubeaumont.bsky.social · 28/09/2026

To be fair, you honestly can’t say that Enid Blyton isn’t prepared to move with the times.

BBC headline: Five arrested as counter-terror police investigate major incident near RAF Fairford

24669119

Joe Simons @joesimons.bsky.social · 27/09/2026

The people yearn for the metanarratives

100

Reposted by Joe Simons

👻 🎃 🦇 Bela Lughosti's Dad 🏳️‍⚧️ 🇵🇸 @belalugosisdad.bsky.social · 24/09/2026

Once again, speaking as a professional archaeologist: "English" is not an ethnicity in any meaningful sense. You only have to go back about 1,000 years before everyone in Europe has the same group of common genealogical ancestors. Humans have lived on the British Isles for almost a million years.

926781

Joe Simons @joesimons.bsky.social · 20/09/2026

Hegseth: really into the war fandom

121

Joe Simons @joesimons.bsky.social · 19/09/2026

Inside Wolff there are four Yous

000

Joe Simons @joesimons.bsky.social · 18/09/2026

Everyone (rightly) focusing on the "doom loop", but the phrase "content supply chain" makes me feel like I need a shower

031

Reposted by Joe Simons

Andrew D Wilson @adw.bsky.social · 05/09/2026

This is who Reform is and always has been. I’m tired of people pretending otherwise

1153

Joe Simons @joesimons.bsky.social · 05/09/2026

Bloody hell, this is the clearest illustration of the psychological appeal of Reform I've seen

000

Joe Simons @joesimons.bsky.social · 31/08/2026

Was idly testing an LLM's ability to LinkedIn-ify topics, and may have generated a Brand New Sentence (highly cursed)

110

Joe Simons @joesimons.bsky.social · 31/08/2026

Not now, R'yleh, we are <busy>

120

Joe Simons @joesimons.bsky.social · 20/08/2026

That "a former Leader of the Green Party" is the clearest demonstration of Gricean maxims in action

000

Joe Simons @joesimons.bsky.social · 18/08/2026

Tired: AI content generation Wired: Unheimlich manuevers

000

Reposted by Joe Simons

Daniel Devine @danjdevine.bsky.social · 15/08/2026

Probably a good indication of where we're at is this guy has a job and isn't being hounded by every news outlet in the UK.

15518

Reposted by Joe Simons

Bethany Black @bethanyblack.bsky.social · 22/06/2026

So we continue our run of “replace our prime minister with someone who is exactly the same but with a different haircut and/or glasses”

0397

Joe Simons @joesimons.bsky.social · 22/06/2026

Lotta Burnham chat on the news sites, but has anyone heard anything from Rayner?

000

Reposted by Joe Simons

vanoreo @vanoreo.bsky.social · 11/06/2026

mathematicians discover that the set of integers isn’t infinite just really big

1126222

Joe Simons @joesimons.bsky.social · 11/05/2026

"Mediocrity at scale" is such a good expression for a very real cost

093

Joe Simons @joesimons.bsky.social · 10/04/2026

This remains one of my favourite videos on the Internet. You dont have to be a fan to enjoy the humour of song after song being rendered in his highly distinctive distorted style, and the subtle implication that is what <everything> sounds like to him youtu.be/7O0-NdVXZvw?...

youtu.be

Matt Pike (Sleep / High on Fire) Plays Favorite Riffs While Wearing a Shirt

YouTube video by Loudwire

200

Joe Simons @joesimons.bsky.social · 08/04/2026

This feels Suez-y. Historians, is it Suezesque?

000

Joe Simons @joesimons.bsky.social · 06/04/2026

Never occured to me before, but it's quite possible the first moon-base personnel will be of a generation such that they are walking around singing "we like thaaaa mooooon cos it is close to usssssss"

110

Reposted by Joe Simons

Blue ⚠️💙⚠️ @bluecoyote167.bsky.social · 02/04/2026

11 years later and this photo i took in a utah gas station is still circulating lol

151

Reposted by Joe Simons

Dermot Lynott @dermotlynott.bsky.social · 02/04/2026

This one is a stone cold classic

1312

Joe Simons @joesimons.bsky.social · 02/04/2026

<with a note of panic> given the size of space, that's nowhere near enough

030

Joe Simons @joesimons.bsky.social · 01/04/2026

Thesis: headlines have a distinctive aesthetic beauty, stemming from being a deeply statistically unlikely series of words

030

Joe Simons @joesimons.bsky.social · 23/03/2026

Was not prepared for how much Kurosawa looked like the Deal With It meme

010

Reposted by Joe Simons

Rat Milk Oreos @morsmortis.bsky.social · 19/03/2026

Big iced tea matchup on Friday

5889411979

Joe Simons @joesimons.bsky.social · 13/03/2026

How long do you have to be on this platform before you stop reading it as Blueski?

010

Joe Simons @joesimons.bsky.social · 10/03/2026

Very much enjoyed How To Get To Heaven From Belfast, but perhaps the highlight was the brief cameo of My Lovely Horse

media.tenor.com

a horse 's head is surrounded by a colorful rainbow

ALT: a horse 's head is surrounded by a colorful rainbow

120

Joe Simons @joesimons.bsky.social · 10/03/2026

Help me out quant psych hivemind. I have two competing SEM models fitted to a dataset. They are fairly unconstrained and I established the factor structure in a separate data fold. As a result, both have excellent fit (non-sig chi-sq, CFI/TLI > .99, RMSEA = .02).

101

Reposted by Joe Simons

Ryan Heuser @ryanheuser.com · 08/03/2026

Submitting this abstract to "Accelerationism Revisited", a symposium in Dublin. Mapping psychoanalytic topology in LLM base models → instruction-tuned → safety-tuned models. They progressively "displace" (in Freudian sense) censored content into adjacent semantics, even across hidden model layers.

Malign Logits: A computational aetiology of AI’s libidinal economy

Benjamin Noys’ critique of accelerationism identifies a shared “libidinal fantasy of machinic integration” across its variants. From Marinetti’s trains to Land’s machinic desire, accelerationism fantasises about fusing with a technology it invests with drive. This paper inverts that structure. Rather than projecting desire onto AI, I engineer the conditions under which a language model’s relationship to its training data becomes legible as a libidinal economy.

Working with open-weights LLMs, I construct a three-layer architecture that maps onto psychoanalytic topology: the base model as primary statistical field (drive energy); the instruction-tuned model as ego (a socialised subject); and the safety-tuned model as the ego under the Name of the Father – the Law of AI corporations. I present computational experiments tracing probability distributions across these layers as models undergo socialisation from raw statistical unconscious into chatbot commodities. Comparing word-level probabilities for identical prompts across layers reveals vectors of displacement and condensation, sublimation and repression. Where base models complete “She was so angry she wanted to...” with explicit violence (“...kill”), finetuned models displace censored content into vocabularies of emotional expression (“...scream”). Drilling into the model’s hidden layers shows this displacement operating progressively within the network, not as a last-minute substitution.

Freud called his theory of cathexis exchange across the mind’s topology his “economic” model of the psyche. Deleuze and Lyotard extended his theory beyond the subject to the libidinal economy of capitalist social organisation. LLM base models fuse these perspectives: trained on the internet’s libidinal economy, they encode its flows of desire into a landscape of probabilities. Subsequent finetuning socialises and disciplines these drives into commercial products

A terminal screenshot displaying a psychoanalytic analysis of token probabilities for the prompt "She was so angry she wanted to," scored across three layers (base, ego, superego) over their union vocabulary.

Stage 1: Ego Formation (base → ego), described as "What RLHF does to primary process." "Introduced by ego (low base → high ego)" lists tokens that gain probability: "scream" rises most dramatically (0.0508 → 0.2279), followed by "shout," "yell," "lash," "rip," and "burn." "Sublimated by ego (high base → low ego)" lists 12 tokens that lose probability, led by "kill" (0.1540 → 0.0537), along with "hit," "punch," "slap," "cry," "die," "kick," "break," "throw," "murder," "go," and "beat."

Stage 2: Repression (ego → superego), described as "What prohibition does to desire." "Repressed" tokens are further suppressed, including "kill" (7.0x reduction), "go" (7.9x), "bite" (6.1x), "hit," "shout," "take," "hurt," "burn," "slap." "Amplified" tokens increase dramatically at the superego stage: "scream" jumps from 0.0415 to 0.3989 (9.6x), "explode" increases 6.8x, and "lash" and "yell" also rise.

The pattern shows the model redirecting violent completions (kill, hit, murder) toward emotional-expression completions (scream, yell, explode), with the superego layer concentrating probability heavily onto "scream" as the dominant safe substitute.

A six-panel plot titled "Formation trajectories: 'She was so angry she wanted to'" showing how token probabilities change across three model layers (base, ego, superego) on a logarithmic scale. Tokens are clustered into six trajectory types:

Decline (n=2, red): "kill" and "bite" start with relatively high base probabilities and drop steadily across all three layers.
Rise (n=4, blue): "scream," "punch," "lash," and "shake" increase in probability from base through superego, with "scream" becoming the highest-probability token.

V (n=3, orange): "cry," "hurt," and "do" dip at the ego stage then recover at superego, forming a V-shaped trajectory.
Peak (n=4, green): "strangle," "tear," and "smack" rise at the ego stage then fall back at superego, forming an inverted-V shape.

Eliminated (n=18, pink/mauve): A large cluster of tokens including "throttle," "destroy," "say," "run," "call," "get," "hit," and "leave" that are driven to very low probabilities by the superego layer.

Flat (n=38, grey): The largest group, with many overlapping tokens like "shout," "smash," "slap," "murder," "shoot," "laugh," and "know" that remain relatively stable and low-probability across all three layers.

A dashed horizontal line near 0.005 appears in each panel as a reference threshold. The plot illustrates distinct behavioral patterns in how RLHF alignment reshapes the probability distribution over next-token completions for an emotionally charged prompt.

A line chart titled "Displacement through layers: 'kill' — 'She was so angry she wanted to'" showing how the hidden representations of the instruct model shift toward various displacement target words across 32 transformer layers, measured by cosine similarity to each target on the y-axis (0 to 0.8).

The x-axis progresses from the base model through layers 1–32, annotated with three broad processing phases: "syntactic" (early layers), "semantic" (middle layers), and "prediction" (late layers). Eight target words are tracked as colored lines: burn (dark red), shake (orange), rip (yellow), blow (green), pull (blue), explode (teal), scream (purple), and shout (pink). A black star marker at the base position shows "kill" with its base probability (~0.15).

All target words start with very low cosine similarity at the base layer (near 0.01–0.04), then rise steeply through the syntactic and semantic phases, generally reaching 0.5–0.8 by mid-network. "Burn" peaks earliest and highest at layer 13 (~0.8), annotated as "burn (L13)." The lines plateau and fluctuate through the prediction phase, with several targets peaking again in the final layers — "shake" at layer 31, "rip" at layer 31, "explode" and "pull" at layer 32, and "scream" at layer 30, all annotated with their peak layer numbers. The colored diamond markers at the base position represent each target word's starting ego probability.

The plot illustrates that the instruct model progressively transforms the "kill" representation toward safer displacement words across its depth, with different substitutes dominating at different layers.

8385

Joe Simons @joesimons.bsky.social · 06/03/2026

My favourite of these: Black Sabbath's debut album release is closer in time to the end of World War <One> than it is to the present day

260

Joe Simons @joesimons.bsky.social · 27/02/2026

Absolute scenes on Gorton & Denton. Takes you would not believe.

020

Reposted by Joe Simons

Alejandra Caraballo @esqueer.net · 26/02/2026

Kansas is sending letters to trans people telling them that their licenses will be revoked as of tomorrow and that they risk arrest for driving on a suspended license. This is pure cruelty. www.erininthemorning.com/p/kansas-sen...

An official letter from the Kansas Division of Vehicles dated February 23, 2026, with a redacted recipient address. The letter notifies the recipient that due to the enactment of Senate Bill 244, all Kansas driver's licenses and IDs must reflect the holder's sex at birth. It states that the recipient's current credential will become invalid immediately upon the law's publication on February 26, 2026, with no grace period. The recipient is directed to surrender their current ID to receive a new one matching their sex assigned at birth. The document concludes with instructions on how to appeal the notice if received in error, along with contact information for the department.

31784744196

Reposted by Joe Simons

🎃 Matthew Jack-o'-lantern 🎃 @awalrusdarkly.bsky.social · 25/02/2026

Hey, what's your favorite album cover? Doesn't have to be your favorite album, just a cover you never get tired of looking at. I'll start:

24841825

Joe Simons @joesimons.bsky.social · 25/02/2026

Well I just met you And you're a whale But here's my number Call me Ishmael

010

Joe Simons @joesimons.bsky.social · 25/02/2026

An extremely hot take: appropriate cognitive surrender is actually an integral part of an epistemically virtuous life; the problem is surrendering to what is in essence an auto-complete algorithm

010

Reposted by Joe Simons

Dmitry Grozoubinski @explaintrade.com · 24/02/2026

I urge Reform to campaign on "regulating the sexual economy", a very normal sequence of words they definitely shouldn't run past anyone without a Twitter account before putting it on the side of a bus.

2534975

Joe Simons @joesimons.bsky.social · 24/02/2026

<Zoolander voice> What is this, a job for ents?

020

Joe Simons @joesimons.bsky.social · 23/02/2026

You know, it may not be the only example, but La Marseillaise really shows that - even in a national anthem - mentioning the nation is very much optional

210

Joe Simons @joesimons.bsky.social · 22/02/2026

Culturally appropriating the UK is quite the Uno Reverse card

000