This was literally a bit on Adult Swim 10 years ago. youtu.be/DJklHwoYgBQ?...
19.10.2025 18:37 โ ๐ 24 ๐ 2 ๐ฌ 0 ๐ 0@danielfennelly.com.bsky.social
Programmer, cat dad, synthesizer knob twiddler, AI experimentalist
This was literally a bit on Adult Swim 10 years ago. youtu.be/DJklHwoYgBQ?...
19.10.2025 18:37 โ ๐ 24 ๐ 2 ๐ฌ 0 ๐ 0TFW youโre home sick and 80% done with Silksong but the brain fog is too strong to stand a chance against these bosses
24.09.2025 21:10 โ ๐ 0 ๐ 0 ๐ฌ 0 ๐ 0Heโs not exactly the sharpest knife in the drawer so he might not be lying.
19.09.2025 15:31 โ ๐ 1 ๐ 0 ๐ฌ 0 ๐ 0"Don't RSVP, that puts your name on a subpoenable list."
IDK. So does donating? So does putting your name down on a ballot-initiative petition? Frankly, so does walking out of your house with your phone in your pocket.
Get rid of the fascist in your head and live like a citizen of a free country.
The Third Reich didn't last 1,000 years.
Pinochet was ousted with a referendum.
And the US isn't exactly dealing with the smartest, most competent fascists.
Those of you insisting that 2025 is forever need to read a book, touch grass, go to therapy, anything other than trying to make others quit.
Taking the small opportunities for consumer activism where I can find them. Canceled Disney/Hulu. So much for my attempt to watch The X-Files from start to finish as a retreat from contemporary horrors.
18.09.2025 15:23 โ ๐ 1 ๐ 0 ๐ฌ 0 ๐ 0Breaking: The Trevor Project received a stop-work order last night on its contract with the national 988 suicide prevention hotline. The Trump administration is eliminating the option for LGBTQ callers to the hotline to press 3 and connect with someone who specializes in LGBTQ mental health.
18.06.2025 13:59 โ ๐ 7935 ๐ 4385 ๐ฌ 252 ๐ 1304As an optimist about the potential upside of AI, I wish people would stop putting together garbage research like this. The authors don't even acknowledge (or realize) that "better search engine over survey results" is an option. Obvious retroactive justification for someone's AI pet project.
11.06.2025 15:04 โ ๐ 2 ๐ 0 ๐ฌ 0 ๐ 0Little update on Facebook's decision to shut my account.
11.06.2025 09:44 โ ๐ 1279 ๐ 287 ๐ฌ 141 ๐ 87A distinguished penguin in a three-piece suit giving a PowerPoint presentation to a boardroom full of rubber ducks about the quarterly reports on why pineapple belongs on pizza, while a confused-looking houseplant takes notes in the corner
A moment of psychocybernetic whimsy. Frontier AI models are (biased) compressions of the collective unconscious of the internet, so I asked Claude 4 to write me a prompt for a playful shitpost which I fed in to Flux 1.1 pro. #ai-generated
11.06.2025 14:41 โ ๐ 0 ๐ 0 ๐ฌ 0 ๐ 0Iโve never forgotten this picture from 2016 and how it completely reshaped my understanding of how the media covers protests.
11.06.2025 13:06 โ ๐ 12261 ๐ 4695 ๐ฌ 97 ๐ 45A jumping off point i generated because itโs been awhile since I last studied these thingsโฆ
www.perplexity.ai/search/945d8...
Setting aside technology for a moment, the relationship between evolution, learning, and culture is fascinating. Learning capacity can reduce pressure on bio evo, but learning and culture can also increase complexity of environment. Sci Show episode idea?
08.06.2025 17:24 โ ๐ 0 ๐ 0 ๐ฌ 1 ๐ 0Using a LLM to write a term paper is useful in a similar way t using a forklift for weightlifting at the gym. Yes, it may be efficient in completing the task, but the value is in the hard work
05.05.2025 16:03 โ ๐ 622 ๐ 149 ๐ฌ 20 ๐ 16Morning chatter on AI alignment spurred a bit of ironic and lazy cybernetic philosophy. I think Rortyโs contingent liberalism could be an interesting starting point for an alignment research program.
www.perplexity.ai/search/9ab37...
Disappointed the article doesnโt distinguish between AI systems that have data augmentation and LLMs which donโt. Not a fanboy but OAIโs mentioned hallucinations matter a lot less than Google and Bing in this context.
05.05.2025 15:19 โ ๐ 0 ๐ 0 ๐ฌ 1 ๐ 0Iโm more optimistic about the use of AI for inter-human alignment than about solving AI alignment in any general fashion. Universal alignment isnโt needed to facilitate compassionate discourse.
05.05.2025 14:43 โ ๐ 3 ๐ 0 ๐ฌ 1 ๐ 0Making progress in post-modern ethics is far more difficult than incremental improvements to algorithms. Generative AI cranks up the volume on unsolved cultural problems.
05.05.2025 14:11 โ ๐ 4 ๐ 0 ๐ฌ 2 ๐ 0At this point, censorship will just be privatized. It will side-step the first amendment worries quite neatly. Weโre certainly in the right frame for it to happen, and distracted enough to not oppose it coherently.
This is the human alignment problem, if you like.
New favorite incantation. First prepare your spellbook.
```
git config --global alias.recent-branches '!f() { count="${1:-5}"; git for-each-ref --sort=-committerdate --count="$count" refs/heads/ --format=" %(refname:short) (%(committerdate:relative))"; }; f'
```
Now:
`git recent-branches 10`
Stupid computer game with a dialogue pop-up needlessly explaining the tutorial popups.
Is #shitvibing a thing yet? You just vibe code your way to a shitpost of a computer program. Earlier I made a game which is just clicking through tutorial dialogues. That's the game.
06.04.2025 06:03 โ ๐ 0 ๐ 0 ๐ฌ 0 ๐ 0Today in software engineering, a website I needed to use was broken. As it loaded I saw it show and then remove a button I needed. I opened the web developer tools and throttled my connection. The website worked afterwards and I shared a good laugh with my coworker. ๐ฅฒ
25.02.2025 01:53 โ ๐ 0 ๐ 0 ๐ฌ 0 ๐ 0It's nearly 5am and Republicans just rammed through their pro-billionaire, anti-middle class budget blueprint.
In order to pass massive tax giveaways for billionaires, they're going to defund Medicaid, slash veterans benefits, & force kids to go hungry.
Make sure everyone knows it.
"Hochul is dithering about whether to use her powers. AOC, in contrast, exercised her 1st Amendment rights to inform her constituents about their 4th Amendment rights. In doing so, she showed that the administration is tyrannical. She also showed that itโs weak โ if youโre willing to confront it."
21.02.2025 16:20 โ ๐ 8165 ๐ 1739 ๐ฌ 146 ๐ 73Kudos to Mapquest for this brilliant bit of fun marketing
19.02.2025 16:43 โ ๐ 2 ๐ 0 ๐ฌ 0 ๐ 0When I was a Hill staffer in 2009, a Dem trifecta was trying to pass universal healthcare. In response, Senate GOP released a memo detailing all of the tactics they in the minority could use to delay the effort as much as possible. I'm recirculating that 2009 GOP memo here. Take inspiration from it.
02.02.2025 17:33 โ ๐ 46892 ๐ 19580 ๐ฌ 1256 ๐ 1840When it comes to these executive orders, I need more states with Democratic governors and majorities saying, simply, โThat shit is illegal. We ainโt doing that. Sue us,โ and go on with business as usual.
01.02.2025 18:00 โ ๐ 665 ๐ 193 ๐ฌ 15 ๐ 9