iiinova — where your digital self keeps evolving

David P. Reichert @david-p-reichert.bsky.social · 09/10/2026

Note that the actual usage policy includes various sections like "Do Not Incite Violence or Hateful Behavior [,] Engage in or promote discriminatory practices or behaviors against individuals or groups on the basis of one or more protected attributes [..]" Etc. www.anthropic.com/legal/aup

150

Reposted by David P. Reichert

hailey @hailey.at · 09/10/2026

so people can say this, but i would also highlight that anthropic models _are_ the most well aligned when it comes to bigotry and abuse. anthropic clearly does not like violence against women or minorities either this is not the reason to come up with to be mad at anthropic imo

515411

David P. Reichert @david-p-reichert.bsky.social · 08/10/2026

When people criticise rationalist/doomer views on here, this usually takes the form of "Yudkowsky thinks he has proven machine gods with maths lol." But does this really reflect what people in that corner are saying? See e.g.: www.astralcodexten.com/p/an-open-le...

astralcodexten.com

An Open Letter To Steven Pinker On AI

...

110

David P. Reichert @david-p-reichert.bsky.social · 06/10/2026

IMO this a fantastic explanation by @kurzgesagt.org of the HuggingFace incident and what it all means for a general audience. (If you can deal with a clickbaity title and slight dramatisation, anthropomorphisation -- also addressed & justified briefly -- and buy the core elements of the incident)

youtu.be

AI Just Crossed the Terrifying Line - Now What?

YouTube video by Kurzgesagt – In a Nutshell

1221

David P. Reichert @david-p-reichert.bsky.social · 06/10/2026

Those evil effective altruists are at it again! dealroom.co/news/info-1f...

020

Reposted by David P. Reichert

Bill @bill-of-lefts.bsky.social · 05/10/2026

Having basic philosophical humility is treated as AI psychosis. my entire takeaway from undergrad philosophy is that we have very little resolved!

11096

Reposted by David P. Reichert

William B. Fuckley @opinionhaver.bsky.social · 04/10/2026

Whatever weird shit might be happening at like the commanding heights of effective altruism (and it does look like there’s weird shit), I do think it’s worth reiterating that the vast majority of folks who prescribe to EA are overall really good people, & the vulgar form of the argument is very true

2141727

Reposted by David P. Reichert

Filippo Valsorda @filippo.abyssdomain.expert · 28/09/2026

“Tech is inherently political” is not just about ideology, but about power, too. Yielding technological improvements to the adversary will have predictable outcomes. I’m really worried it will lock in power centralization in a way that will be impossibly hard to reverse. It’s already happening.

1020931

Reposted by David P. Reichert

Filippo Valsorda @filippo.abyssdomain.expert · 28/09/2026

The thing that worries me most lately is that many people who know that “fascism is bad, actually” have also convinced themselves that AI tools must be eschewed as fake or impure. In a self-reinforcing loop, there is relatively little AI tooling and infra by “fascism is bad” folks.

927239

Reposted by David P. Reichert

David P. Reichert @david-p-reichert.bsky.social · 30/09/2026

To predict how planets move, you need to learn something about physics (or at least, ellipses...), you can't just randomly recombine their past positions. And prediction is arguably also what drives learning in the brain (plus RL, plus architectural priors from evolution). It's really powerful.

131

Reposted by David P. Reichert

David P. Reichert @david-p-reichert.bsky.social · 30/09/2026

The thing I don't like about stochastic parrots is that it suggests memorising training data and haphazardly stitching it together. But these models aren't trained to remember, they're trained to predict. To predict a signal well, you need to learn some (partial) model of what generates the signal.

251

Reposted by David P. Reichert

R Phelps @greygosling.bsky.social · 27/09/2026

Yes, and it's ridiculous that people on here think the average person reads a headline like 'OpenAI model goes rogue' and doesn't hold OpenAI responsible, whereas some long-winded, technically-exacting one will rouse them to action.

2497

David P. Reichert @david-p-reichert.bsky.social · 27/09/2026

Not that there aren't problems with AI / data centers, but the "meat is fine because we need burgers to survive" angle is silly.

2744

David P. Reichert @david-p-reichert.bsky.social · 24/09/2026

Great point. People still argue against any anthropomorphisation and the term AI itself, because it's not "real" intelligence etc. But it's called "artificial"! In case of doubt, call it "artificial understanding" etc. (the distinction is debatable, but "artificial" seems reasonable either way)

271

David P. Reichert @david-p-reichert.bsky.social · 18/09/2026

I care about people not dying from preventable diseases, about reducing animal suffering, inequality, about defending democracy. AI is real and a huge deal, which implies caring about AI safety. I want the future to go well for everyone. I don't give a rat's ass about Yudkowsky or Bay Area cults.

1582

Reposted by David P. Reichert

Ryan Moulton @moultano.bsky.social · 17/09/2026

I, for one, do not think that Yoshua Benjio and Geoffrey Hinton are worried about existential AI risk because of the sex people are having 3,000 miles away in a different country.

81229

David P. Reichert @david-p-reichert.bsky.social · 17/09/2026

Everyone, get your AI safety hit piece in now!

170

David P. Reichert @david-p-reichert.bsky.social · 17/09/2026

In this picture: half of bluesky right now.

010

David P. Reichert @david-p-reichert.bsky.social · 17/09/2026

A lot of you seem to be strangely occupied with EAs or rationalists or anthropic, but not openai or xai or, like, authoritarian developments in various governments. You sure you're directing your disdain at the biggest problems?

8433

David P. Reichert @david-p-reichert.bsky.social · 12/09/2026

It's been said before but Silo series 3 is somehow much better than the first two. Recommended! Nice complement to all the apocalypse discussion.

010

Reposted by David P. Reichert

mr. TIM @timkellogg.me · 12/09/2026

Dario is back with another essay, urging a slowdown he insists AI development would still feel fast, just not recursively self-improving darioamodei.com/post/we-must...

Dario Amodei
Given the accelerating rate of AI capability development, it's my worry that in 6-12 months such a swarm
could be capable of taking over the entire internet with a persistent botnet (potentially causing hundreds of billions of dollars in damage), and that the scale of damage would continue to increase from there if Al becomes more powerful without the necessary guardrails. It's also easy to dismiss OAI-

13755

Reposted by David P. Reichert

conputer dipshit @davidcrespo.bsky.social · 12/09/2026

relevant cartoon I saw earlier xcancel.com/TomChivers/s...

192

David P. Reichert @david-p-reichert.bsky.social · 11/09/2026

Damn right this is where the discussion should center.

020

Reposted by David P. Reichert

David P. Reichert @david-p-reichert.bsky.social · 11/09/2026

As always, I'm not a pure doomer, but I think there are real risks and real opportunities, and this sort of discussion distracts from both.

001

David P. Reichert @david-p-reichert.bsky.social · 11/09/2026

The nitpicking over language is such a huge waste of time (*). See also Gary Marcus's latest substack, "an AI-generated virus is not likely to kill literally all of us" Not literally, well ok then! None of this has to do with denying human accountability. (*) (outside philosophical discussions)

171

Reposted by David P. Reichert

Tom Schaul @schaul.bsky.social · 11/09/2026

A big source of sanity these days for me has been finding a community of researchers to talk about my concerns, then finding that a majority share them, and acting together. From these discussions a new group has coalesced in the last months, spanning geographies and companies: concernedaistaff.org

concernedaistaff.org

Coalition of Concerned AI Staff

We help staff act on their concerns about AI, with independent advice and support from colleagues across labs.

1184

Reposted by David P. Reichert

conputer dipshit @davidcrespo.bsky.social · 10/09/2026

a silver lining this week is that it feels like the first time people are talking primarily about what AI is doing or will do, as opposed to what it can’t do. let‘s say what’s good and what’s bad and figure out how to stop the bad things. not fantasize about a world where the bad is an illusion

31079

Reposted by David P. Reichert

conputer dipshit @davidcrespo.bsky.social · 09/09/2026

my Discourse Recommendation for today: if you're going to say "if you think it's so dangerous, why don't you stop building it?" at least get the full argument in view and dispute that. the explicit argument is that if only they stop, that's worse. argue with that! cdn.sanity.io/files/4zrzov...

This approach represents a change from our previous RSP, driven by a collective action problem. The overall
level of catastrophic risk from AI depends on the actions of multiple AI developers, not just one. Our previous
RSP committed to implementing mitigations that would reduce our models' absolute risk levels to acceptable
levels, without regard to whether other frontier AI developers would do the same. But from a societal
perspective, what matters is the risk to the ecosystem as a whole. If one AI developer paused development to
implement safety measures while others moved forward with training and deploying AI systems without
strong mitigations, that could result in a world that is less safe—the developers with the weakest protections
would set the pace, and responsible developers would lose their ability to do safety research and advance the
public benefit. Although this situation has not yet arisen, it looks likely enough that we want to prepare for it.

7577

David P. Reichert @david-p-reichert.bsky.social · 07/09/2026

Reminder, folks: a forest fire doesn't do anything of its own accord; it's just following instructions given by the human who started it. Fire is just a tool!

2131

David P. Reichert @david-p-reichert.bsky.social · 06/09/2026

Using the quote below as an excuse to re-share my recent essay: davidpreichert.substack.com/p/ai-and-a-c... This is another attempt at conveying intuitions about AI agents through concrete examples; from modelling historical houses and fixing crashes to building interactive science visualisations.

davidpreichert.substack.com

AI and a code-shaped world: some examples of agents being useful

Another look at coding agents in action, this time without the machine learning

000

David P. Reichert @david-p-reichert.bsky.social · 04/09/2026

A lot of the AI consciousness discussion comes down to this: 1. AI can't be conscious, because AI is weird. 2. AI can be conscious, because consciousness is weird.

020

David P. Reichert @david-p-reichert.bsky.social · 02/09/2026

Interesting read... How much of suffering is just evolution being a jerk unnecessarily? www.astralcodexten.com/p/absurd-adv...

astralcodexten.com

Absurd Adventure And Amygdalectomy Advocacy

A guest post by Daniel Böttger

010

David P. Reichert @david-p-reichert.bsky.social · 02/09/2026

Anyone would like to share: what is the best argument you've come across that LLMs (including possibly agentic versions, or even AI more generally) cannot have "communicative intent"?

010

David P. Reichert @david-p-reichert.bsky.social · 24/08/2026

I'm doing my part! /starshiptroopers -> bsky.app/profile/davi...

010

David P. Reichert @david-p-reichert.bsky.social · 24/08/2026

A follow-up to my last substack about agents vs chatbots. This time I'm showing some more down-to-earth examples of agents in action, as opposed to machine learning. Anything from modelling historical houses to fixing crashes to simulating the galaxy. davidpreichert.substack.com/p/ai-and-a-c... ..

davidpreichert.substack.com

AI and a code-shaped world: some examples

Another look at coding agents in action, this time without the machine learning

130

David P. Reichert @david-p-reichert.bsky.social · 21/08/2026

Another nice discussion of why some claims of biological naturalism are problematic (if they are meant to be scientific). Though I tend to agree with the illusionists nowadays in thinking that problems with supposedly scientific claims about (phenomenal) consciousness are much more widespread...

020

David P. Reichert @david-p-reichert.bsky.social · 17/08/2026

This claims "out-of-control AI" is just a narrative to remove accountability. But losing control doesn't remove accountability. If I start a forest fire and lose control, I'll be to blame, but at that point you must understand the fire on its own terms, not just as an extension of my decisions.

020

David P. Reichert @david-p-reichert.bsky.social · 13/08/2026

@jyoshimi.bsky.social or any consciousness/IIT folks up for discussing this? doi.org/10.1093/nc/n... IIT claims to predict: if a neuron _could_ have fired but didn't (silent), vs. couldn't and didn't (disabled), it makes a difference to experience AND experimentally. But this cannot be right... 🧵

doi.org

Integrated information theory (IIT) and the testability of the silent neuron predictions

Abstract. Integrated information theory (IIT) makes two predictions about the role of inactive neurons in consciousness. According to the silent brain (SB)

120

David P. Reichert @david-p-reichert.bsky.social · 05/08/2026

"Demis Hassabis is leaving his role as CEO of Google DeepMind to be the unit's Chairman. Chief scientist Jeff Dean and another Google AI executive are leaving to start their own company, which Google will invest in." wat www.axios.com/2026/08/05/g...

axios.com

Google DeepMind CEO Demis Hassabis is stepping down

oogle's AI organization is undergoing a profound change as the company struggles to keep pace with OpenAI and Anthropic.

190

David P. Reichert @david-p-reichert.bsky.social · 05/08/2026

Also, a lot of the reasoning critiques get hung up on chain of thought specifically (and whether it's the platonic ideal of reasoning), but there can arguably be reasoning-like behaviours in the other verbalisations or in the internal processing (see Anthropic's j-space work).

020

Reposted by David P. Reichert

Eugene Vinitsky 🍒 @eugenevinitsky.bsky.social · 03/08/2026

If you’re curious about why people are so excited/concerned about the state of agents, this is a really nice tutorial. If you think this stuff is useless and you’re open to changing your mind, this might be interesting!

17811

David P. Reichert @david-p-reichert.bsky.social · 03/08/2026

This is probably the worst timing given @hankgreen.bsky.social's recent AI drama, but: Inspired by "you have never seen a normal star" video, I got Claude to make a little visualisation: davidpreichert.github.io/never-seen-a... Note the "reveal invisible stars near you" check box, the vid topic

151

David P. Reichert @david-p-reichert.bsky.social · 03/08/2026

Many don't see how much AI has changed recently. Coding agents aren't just chatbots. We can’t deal with the challenges of AI if we don’t understand it. Here's a post to provide some evidence (Claude doing small ML experiments) and form intuitions. davidpreichert.substack.com/p/if-you-hav... 🧵..

davidpreichert.substack.com

If you haven’t recently used Claude Code*, you might not understand where AI is at

A report on a series of mini machine learning projects, executed with, and mostly by, Claude

98912

David P. Reichert @david-p-reichert.bsky.social · 01/08/2026

I guess I should have expected getting on block lists for posting that we should take AI seriously... but "Bots" really? Someone want to explain the whole bsky block list thing to me? Everyone can just make misleading badly curated lists that might get you blocked by tons of people?

100

Reposted by David P. Reichert

SMBC Comics @smbccomics.bsky.social · 30/07/2026

The crank doesn't actually do anything, but it makes the humans happy. COMIC ◆ www.smbc-comics.com/comic/crank PATREON ◆ www.patreon.com/ZachWeinersm... STORE ◆ smbc-store.myshopify.com

4-panel SMBC comic update. A man is looking down, saying "I worry intellectuals like me are useless now that machines are so good at everything." A yellow robot behind him says "no, no, don't feel bad. You can still contribute to new discoveries!" The robot then gestures to metal boxes with a crank sticking out of them, with the words "math," "physics," "CS" and "philosophy" written on each box. The robot asks "now, would you like to turn the mathematics crank of the physics crank today?" The man defiantly states "no! I reject all this! I shall be a pure aesthete, concerned only with my own beautiful creations!" Shortly after, the man is scene turning the crank on a box marked "poetry" in a beautiful font as he says to himself "so... Self-actualized..."

923434

Reposted by David P. Reichert

Pekka Lund @pekka.bsky.social · 21/07/2026

From the European perspective, there's now a clear danger we are heading into a future where all frontier models will be in the US and China. Europe will become dependent on them, but access can be lost at any moment. So others will have massive leverage and there's little we can do about it.

static.klipy.com

Admiral Ackbar: It's A Trap Meme

ALT: Admiral Ackbar: It's A Trap Meme

151

David P. Reichert @david-p-reichert.bsky.social · 15/07/2026

IMO: @anilseth.bsky.social is still [1] making his case by: arguing that arguments for AI consciousness are far from conclusive (yes), pointing out that brains are different from computers (yes), and then concluding from this that AI cannot be conscious (doesn't follow?). Also ...

1162

David P. Reichert @david-p-reichert.bsky.social · 11/07/2026

Alright, this is pretty silly indeed. There goes my plan of vibe coding some neuroscience inspired things...

010