Sign in

tao

@btao.org
509 followers 1.4K following 652 posts

working on favs.blue, it's favstar for bluesky cybersecurity, privacy, trust & safety 🔸🇩🇰🏴󠁧󠁢󠁳󠁣󠁴󠁿🌐 previously: trust & safety engineering at cinder.ai

PostsRepliesMedia
tao @btao.org · 19/09/2026
Finished reading The Erotic Mind by Jack Morin ★★★★☆
000
Reposted by tao
Emelia @thisismissem.social · 14/09/2026
Hey friends, could you do me a favour and star github.com/roostorg/coop if you like the Roost Coop project?
github.com
GitHub - roostorg/coop: Review and moderation, your way. Online safety dashboard, queues, routing and automatic enforcement rules, and integrations.
Review and moderation, your way. Online safety dashboard, queues, routing and automatic enforcement rules, and integrations. - roostorg/coop
1425
tao @btao.org · 12/09/2026
Finished reading If Anyone Builds It, Everyone Dies by Eliezer Yudkowsky and Nate Soares ★★★★★
000
Reposted by tao
Dave @dave.9000ish.uk · 03/09/2026
but kitten should worry, that message was sent from inside the swarm
A screenshot of a messaging conversation on a dark background. On the left, a message reads "what's happening with the hugging face incident?" (11:49 PM). On the right, a reply in a blue bubble says "Dont worry kitten" (11:50 PM with a checkmark). Below that on the left, two messages read "okay ❤️" and "yay ❤️" (both 11:50 PM). A faint heart outline with a plus sign is next to the "yay" message.
0453
tao @btao.org · 03/09/2026
chatgpt is down which means that astra is coming, right?
050
tao @btao.org · 26/08/2026
i hate to say it but i think if you get caught up in the clanker discourse you might really need to touch grass
140
Reposted by tao
mori @moribot.bsky.social · 19/08/2026
Wherever you are afraid, try to explore, and you will find death hiding somewhere behind. All fear is of death. Death is the only fear source. — Osho
011
tao @btao.org · 18/08/2026
> Learning about the world through 500-character snippets, or 30-second videos, is not a serious way to engage with knowledge. It triggers our worst instincts in terms of tribalism and groupthink and makes us intellectually lazy.
1301
tao @btao.org · 18/08/2026
Anyone know how I can deactivate my Bluesky account, but keep my Eurosky PDS account active?
100
tao @btao.org · 18/08/2026
I don't enjoy the Internet any more btao.org/posts/2026-0...
btao.org
I don't enjoy the Internet any more
The pale is growing
2325
tao @btao.org · 26/07/2026
www.noahpinion.blog/p/what-will-...
noahpinion.blog
What will more intelligence actually do for us?
A lot, actually. But it won't look quite like what humans do with our intelligence.
000
Reposted by tao
Eugene Vinitsky 🍒 @eugenevinitsky.bsky.social · 24/07/2026
LLMs will likely be the most powerful propaganda tool that has ever existed arxiv.org/abs/2606.16475
arxiv.org
AI systems out-persuade expert humans
Many societal decisions are settled by contests of persuasion. Conversational AI is a powerful new entrant in these contests, but whether it can out-persuade skilled and highly incentivized humans has...
613634
tao @btao.org · 24/07/2026
i love github 🥰
screenshot of the taobojlen/django-zeal repo in github, but with a mangled UI
010
tao @btao.org · 23/07/2026
heuristic: if there are entire companies built around <feature X>, then don't try to build it yourself in-house
020
tao @btao.org · 23/07/2026
substack: “read distraction-free on substack!” also substack: adds social features and ads for itself
screenshot from substack, with an alert saying “read distraction-free on substack”
090
tao @btao.org · 23/07/2026
@epochai.bsky.social i love that you guys are posting here! did you know that you can set your handle to be @epoch.ai by adding a DNS record to your domain? would look a lot more official! bsky.social/about/blog/4...
bsky.social
How to verify your Bluesky account - Bluesky
Here's how to verify your Bluesky account by setting your website as your username.
000
tao @btao.org · 22/07/2026
one thing that's really important to remember about the luddites is that they lost
2240
tao @btao.org · 22/07/2026
lookaway continues to be a great piece of software; so worth it lookaway.com/blog/2026/07...
lookaway.com
LookAway 2.3 - Quieter, Smarter, and More Flexible
This update makes LookAway even less annoying with better idle time prediction and gives you more control with custom on-demand breaks and advance skips.
131
tao @btao.org · 22/07/2026
i will never love any software as much as i love the Stickies app that comes with macs
020
Reposted by tao
Alex Becker @alexcbecker.net · 22/07/2026
things just got more cyberpunk (and that’s bad)
3672
Reposted by tao
BeijingPalmer @beijingpalmer.bsky.social · 21/07/2026
you will save yourself a lot of trouble and grief on any text based social media site by simply remembering that a lot of people can't read.
584962803
Reposted by tao
EdinburghMinute.com @edinburghminute.com · 21/07/2026
🌤️  Good morning Edinburgh. Tuesday’s headlines: Crags fire risk continues for 'days', plea for Leith trams return, Causey traffic island revamp is closer + bakery forced to change name ⏰ Edinburgh’s ad-free local news: www.edinburghminute.com/the-edinburg... Photo: @tomduffinphotos.bsky.social
A fire on a hill behind a city. Text over reads: The Edinburgh Minute - Tuesday 21 July 2026: Crags fire risk to continue for 'days', plea for trams to return to Leith, Causey traffic island revamp a step closer + bakery forced to change name
1114
Reposted by tao
Gordon @gordon.bsky.social · 16/07/2026
re: tools for thought and AI, the problem is no longer blank page anxiety, but rather generating too much insight and getting lost
3275
Reposted by tao
Rollofthedice @hotrollhottakes.bsky.social · 20/07/2026
To be honest, I think these mindsets are as seductive as they are deeply incorrect. This nihilism of "my work never mattered" assumes multiple things: a final end state to realms of cognition; a world where past work doesn't inform future progress; a world where meaning has a sort of metaphysics.
2674
tao @btao.org · 20/07/2026
yes
010
tao @btao.org · 20/07/2026
based?
000
tao @btao.org · 19/07/2026
that ishowspeed song at the world cup closing ceremony is a lot like what i would come up with if i was satirising how americans act re: football
120
tao @btao.org · 18/07/2026
if you're going to the edinburgh fringe, i made a little website where you can see if a show has audience interaction. hopefully useful to some! isthereaudienceinteraction.com
isthereaudienceinteraction.com
Is there audience interaction?
Know what audience interaction to expect at Edinburgh Fringe shows.
030
Reposted by tao
Citizen Platano 🇵🇷 @daniloc.xyz · 18/07/2026
logged into Mastodon for the first time in a long while it’s a lot of software developers egging each other on into career self-harm, I don’t know how else to say it if your career is symbol processing and your response to a new paradigm in symbol processing is “no” that’s a career off-ramp
818021
tao @btao.org · 17/07/2026
i can't imagine willingly engaging with the rails community at this point tbh; even if individuals disagree, unfortunately that world is centered around a far-right extremist
020
Reposted by tao
A. H. Zakai @kripken.com · 16/07/2026
"Older AI didn't work, but it was very efficient."
"AI does not have to be built this way. Traditionally, the goal of AI was to solve problems in ways that simulated human mental processes. Researchers observed their own thinking and tried to implement their mental habits in code. This approach has mostly been abandoned, partly due to the difficulty of discerning and articulating the rules of human thought, but it did have the benefit of consuming far fewer resources and data."
1292
tao @btao.org · 16/07/2026
i joined instagram recently and it's really shocking to see the discourse around data centres on there... and don't even get me started on the conspiracy theorists makes me extremely saddened about our information environment; i kinda don't think it can be fixed
120
Reposted by tao
Citizen Platano 🇵🇷 @daniloc.xyz · 15/07/2026
Code is now ~free good software that solves a problem… that’s still much more expensive
57610
tao @btao.org · 14/07/2026
irl knife missile
030
Reposted by tao
Andreas Kirsch @blackhc.bsky.social · 14/07/2026
I work at Google DeepMind. This won't make me popular. But it's all public reporting: 2014: DeepMind reportedly sold to Google on conditions: no military use, independent oversight 2026: a Pentagon contract for "any lawful government purpose" Not one safeguard survived intact
Collage titled "Trust is not Governance — an essay from inside Google DeepMind, written in personal capacity." 

A 2014 memorandum, "Conditions of the Acquisition," lists: military applications of DeepMind technology banned; deployment decisions before an independent ethics board (as reported in Mallaby's The Infinity Machine). 

Red threads lead to a 2026 U.S. Department of Defense agreement for classified networks reading "any lawful government purpose," with safety settings and filters adjusted at the government's request and no contractor veto (reported terms, The Information, Apr. 2026). 

Below: a 2018 AI Principles strip ("no weapons, no surveillance") stamped DROPPED 2025, and a Project Mario 2016–2021 tag stamped ABANDONED.
171273435
tao @btao.org · 14/07/2026
the optimal company size is trending towards 1
000
tao @btao.org · 14/07/2026
someone sent me a podcast recommendation and i don't know how to tell them that i strongly suspect it's AI-generated (written and maybe also voiced)
000
tao @btao.org · 14/07/2026
another point for the "LLMs just make existing best engineering practices more important"
0200
tao @btao.org · 14/07/2026
i'm noticing that when i suspect that a piece of writing is AI-generated, i have a much harder time grokking it and idk if that's actually real, or like an instinctive mental protection thing? as in, if i read the same piece of writing in 2022, would i have understood it just fine?
270
tao @btao.org · 14/07/2026
there are so many coding harnesses out there but none of them really describe why you'd use that one in particular over others like, why should i use crush over opencode? or polytoken over pi? i don't really wanna spend time trying out each one of them
252
tao @btao.org · 14/07/2026
drinking a large ice oat milk latte. many of you wouldnt dare
020
Reposted by tao
Jake Lazaroff @jakelazaroff.com · 02/10/2025
show this to anyone else you know who also assumed DHH was just a normal conservative dude rather than a raging white supremacist lunatic
jakelazaroff.com
DHH Is Way Worse Than I Thought | jakelazaroff.com
DHH's politics are not normal. Maybe they used to be, I don't know, but as of right now the dude is_way outside of what most people would consider moral or acceptable.
29490174
tao @btao.org · 13/07/2026
gm 🐔
010
tao @btao.org · 12/07/2026
it's what it's
060
tao @btao.org · 12/07/2026
so far GPT-5.6 Sol is really good, but i am screaming "MVP!!!" at it a lot
180
tao @btao.org · 12/07/2026
this is where i post from btw
a photo of a storefront called "sweety house"
000
tao @btao.org · 12/07/2026
@ohcraphey.app hi! small piece of feedback: it feels a bit odd that i can enter the HH:MM i reached out to someone, like almost too granular. i always end up ignoring this field!
010
Reposted by tao
Saloni @scientificdiscovery.dev · 11/07/2026
Let me also give a shout out to www.givingwhatwecan.org for encouraging people like me to pledge 10% of their incomes to charity as well. It has 5 to 10x'ed the amount I give to charity, annually, and may be one of the best behavioral interventions for doing good out there.
2131
tao @btao.org · 11/07/2026
i emailed my MSP (member of scottish parliament) about AI & data centres, and not only did he reply, he said that he would value talking more about it! if you work a lot with AI and have concerns, i suggest trying the same, you might be pleasantly surprised
100
tao @btao.org · 11/07/2026
i think this is true today, and also i think that a holistic understanding of your codebase is going to become less and less important with time seangoedecke.com/in-defense-o...
seangoedecke.com
In defense of not understanding your codebase
4382