AI can relieve toil, but it cannot replace the human work that makes a church a community. Silicon Spiritualities season 1 finale: YouTube: www.youtube.com/watch?v=1fWY... Spotify: open.spotify.com/episode/4Gu6... Apple: podcasts.apple.com/us/podcast/i...
That funny feeling when the first talk of the day at the seminar I'm at had to be restructed because the big motivating conjecture has been oneshot by OpenAI. Funny here is bad.
ngl it's undeniable at this point that AI will have pretty significant disruptions in many areas of the economy, people really show a lack of equilibrium thinking when it comes to labor markets. We may have far less mathematicians in the future, but math as a concept hasn't been totally solved lol.
There's a difference between what's for machines and what's for humans, and we shouldn't juice our own productivity by wasting other people's time. I use AI on most of what I write, usually: rough notes in my own voice, iterate, edit, repeat. My blog posts are taking more time, not less, because the feedback loop is fast enough that I go round it more often instead of rereading the same draft and losing sight of how it can be better. Is it making my writing better? Sometimes. Sometimes it's helping me get something into words sooner. But one thing I'm fanatical about is information density. @jnsgruk on going back to writing it himself: jnsgr.uk/2026/10/write-it-your…
the implications of the ease it takes to reverse engineer stuff using ai programming tools has me really really spooked about a future where the majority of software is streamed
The talk of LLM deskilling has me thinking - is "I can't write correct switch statement syntax without IDE autocomplete" an example of IDE deskilling and a black mark against my abilities as a developer? 🤔
Who knew generative AI would have an upside for real artists? Trying to explain to then that I had thousands of ideas, and that it was the writing that was the hard part, was such a struggle. More they're gone, trying to craft the perfect prompt instead of a story. #booksky #indieauthors
the good thing about the flood of AI submissions to conferences is that the acceptance rates are now lower, so the conferences are becoming more prestigious (/s)
I could be barking up the wrong tree, but I feel like I may have stumbled upon why some heavy LLM users go off the rails. (And I’m sure someone else has already suggested this)
I absolutely hate the current environment of massive data collection by corporations and every other institution we interact with. However, polluting the data by purposely supplying wrong information to mess up the algo could seriously backfire. That incorrect data you think is funny could cost you a job, housing, medical care, insurance or other benefits when that info becomes part of the criteria used by the computers to decide who gets denied. It will be difficult to argue incorrect information was used for the determination when they can show you are the one who provided that information. The first round is to run the application through the AI to weed out. Don’t get weeded out because you tried to be funny. There are better ways to fight back. #ai #dataprivacy
If we use AI to “clarify” or add details to a picture, why not use it to “clarify” or add details to a letter or a brief biography written by a deceased relative? Where does one draw the clarification line? #genealogy
I got my first pretty-useless AI/#clanker-driven Security Advisory report on #github. Not that it was _wrong_, just that it's a report for a demo driver in my hobby lzw project (clobbers input file if same as output file), which I only published binaries for to learn how it works. I guess I should actually just remove all releases, since the release flow is semi-broken anyway.
They're using AI to reverse-engineer #adobe apps and releasing them as open-source for FREE and multi-platform. This is the kind of AI stuff I want to see in the art space. This gives power and tools to the artists. #AI
On the whole I agree with the article. If orgs push AI onto their devs, the least they can do is enable them to say "NO!" to a massive 5K+ line PRs of gibberish. If you have to use LLMs for coding, at least spend some time producing code that's easy to review (smaller stacked PRs come to mind here). The job has changed either way however. leaddev.com/career-development…
I've seen two documentaries in the past couple of weeks that employed AI animation. In both instances, they took me out of the movie because it was so distracting and just gave them the feeling of cheapness. On top of that, either could have employed other methods or cut those parts altogether.
Remember the breathless urgency around Mythos? How it was simply “too dangerous to release” and so on? Remember how a lot of folks smelled bullshit and called it a marketing stunt to sell you Claude? Let’s check in on how that is going… pivot-to-ai.com/2026/10/07/myt…
To this day the AI doom crowds pretends like there is even the theoretical possibility of AI safety regulation being a bipartisan issue. it’s just so profoundly unserious
The Anti AI fanatics are soon running out of ways to handle their cognitive dissonance I feel. From an expert on one of the math problems Open AI published solutions to yesterday: "I've seen a lot of clever counting arguments before in graph theory but this is the first time I've seen complex roots and annihilating terms like this, the symbolic manipulation tricks in this look like things out of quantum physics. I don't understand where this trick originated, I need to really digest this." #AI news.ycombinator.com/item?id=4…
It's really ironic how everyone was saying "Play indies and use open source software" right before ai took over every single small project on the internet
"The tech industry won't build this because it turns a 'scalable software' business into a highly disciplined human institution, which scares Wall Street. But from a pure engineering standpoint, it is the only real way to break the algorithmic feedback loop of mediocrity." I proposed to #GoogleGemini on how to flip the paradigm on AI training to create a model that makes fewer errors and hallucinations -- the result of the feedback loop that happens when you keep training on cheap shit. Gemini said Wall Street would hate it even if engineers loved it. :NiceThumb:
there's a metaphor I want to make here but it involves the word "nigger" and folks won't like that
The Italian Brainrot Meme genre may be best known for its absurdist humor, but a lawsuit raises questions about AI, meme culture and the future of IP law. plagiarismtoday.com/2026/10/07… #Copyright #AI #Brainrot #Meme #TungTungTungSahur
i do genuinely recommend emailing kq about your concerns re: ai usage in videos. i did it the last time this issue came up and am doing it again now; it's one of the more formal channels to send in feedback. being loud on socials is totally fair but direct feedback and commentary also helps.
Would #AI exist without the funding from billionaires? I mean without the crazy concentration of wealth where people can invest massive amounts and not fear that losing it will impact their lives, how would modern life be different? Any good idea would still happen, just slowly without causing so much disruption so quickly that it can outpace the regulation we eventually decide it needs. The market story is that people are supposed to slowly grow sustainable businesses and face competition.
With AI & Mathematics (AIM) we are writing a draft vision document that we would like to discuss with the Dutch mathematics community. Please join us: AI4Math Workshop: Shaping the Future of Mathematics with AI Date: Friday 4 December 2026 Time: 09:30 - 15:00 Website: ai4math.nl
the problem I consistently have with antis is that I agree with a lot of the things they are upset about but then they start entering The Cope Zone and saying that because AI can cause bad things for society that it is also useless/a scam.
GPL apps/distros should not be using AI. GPL is not compatible with the ToS of these ai-generating code companies. It was disheartening seeing #Debian going all-AI. Use AI to find security holes, but not to write any code. #ai #artificialintelligence #codex #cursor #programming #gpl #foss #opensource #linux #chatgpt
This is an essential thread to read from a printmaker about printmaking style effects generated by AI and the fact they are disingenuous.
Next was an interesting talk by Sharon Li on using traditional ML methods to improve agent reliability at the Simons Institute youtube.com/live/5vBzKW3-wAQ?i… (3/4)
With you on this. I don't do it with AI or coding so much. I do it with weapons and I integrate things in there and I constantly improve what I'm doing. Take them out test drive them in the field. Make changes until I have the absolutely perfect form of what I want. Then we mass produce them.
“Microsoft is turning to an old friend and a familiar advantage in search of a new edge in AI. The company and its longtime partner Nvidia on Wednesday started taking orders for a new line of high-end Windows PCs built to run large AI models independent of the cloud. Microsoft also announced what it calls “Windows Hybrid Intelligence,” a set of features that will let its Copilot assistant read the files on a PC, take actions on the machine and opt to run AI models locally instead of in the cloud when it makes sense.” geekwire.com/2026/microsoft-pl…
they're tryna call it 'synthography' now
would free credits change your stack? i'd still want the paid bill to make sense. anthropic's new startup offer includes $1,000 in api credits for approved companies. that's room to test a small product, though i'd price the same workload without the subsidy before getting attached. would you experiment or fund what already works? [source](techcrunch.com/2026/10/06/anth…)
I think the AI math proofs are good and cool. Like when MRIs allowed doctors to easily see what before they could only guess at.
Here's my suggestion for overcoming #genAI #LLM burnout and anger: find a project with a strong anti AI stance & contribute to it. My personal suggestions would be 1. Servo (no need to know #rustlang - the book needs your help, too!), 2. Redox (same - docs, end user testing etc are all valuable!), or projects by @glyph I also have a longer list of random nerd snipe topics if none of these tickle you, e.g. Cortex-M backend for cranelift :D 1) servo.org/ 2) redox-os.org/
- @Woot100@void.lgbt
I bet you there's fifteen miles of AIslop behind it these days. Because they think everyone listens to normie music and didn't consider that maybe "guy screaming over five different samples of a trash compactor" would be the sort of situation that requires a human touch. Even before that, though, the second they ended their partnership with Genius Lyrics it was all downhill from there.
‘No productivity increase’ seems fundamentally at odds with my experience but, yes, manually reviewing AI code will of course limit any productivity gains.
I tried to vibe-code a co2 estimator, using a local LLM on my server. But there are several problems: - my Mac is really fast - answers in seconds. The server doesn't have a GPU, and takes over a minute to answer! - the answers are very varied - see screenshot I'll try if the new "Decision models" are better for a task like this.
On-device today, on-chip tomorrow: the model ships inside the boundary you hold. Ownership stops being procurement and becomes the default you never rent. Production-Ready LLM Routing: #buildinpublic #startups https://seldondance.substack.com/p/production-ready-llm-routing
note LLM are good at picking "good mutation" for mutation testing en.wikipedia.org/wiki/Mutation…
The use of AI in medical research is becoming more prominent & harder to detect. "If medical schools & universities don't want the literature to be overwhelmed with garbage then they should probably stop giving every applicant an incentive to publish crap." www.medpagetoday.com/special-repo...
So we've had years of AI coding, billions spent in terms of money and man/computing hours, are there any genuinely great AI coded aps yet? Not pre-existing projects now with AI and new bugs, not AI driven retro game ports built on other people's code for other systems that are completely unoptimised and likely never will be, I mean new thing completely from scratch built by AI that people view as a shining example of what it can do.
You used to be able to google like "I want to know about X" and get a bunch of useful valid search results on the internet and now you get a bunch of random reddit posts, or LLM generate responses that cite random reddit posts
"Frontier models are wedged in a mode where everything they write is a magazine headline. Headlines are good, but you'd wonder about someone who wrote an article with dozens of them." #ai #writing #buzzfeed #slop sockpuppet.org/blog/2026/09/17…
Once upon a time Claude made up some numbers instead of doing actual arithmetic to find an answer and I told it to never do that again, so now if it wants to know if 1+1=2 it spins up a Python script to be sure.
“We do not contend that the AI ‘went rogue,’” plaintiffs’ counsel Rod Cooks of Birmingham, Alabama-based Winston Cooks said via email. “Our concern is that the opposite may be true: an AI system can operate exactly as designed and trained while perpetuating patterns of past discrimination.”
And heading off the drive-bys: Yes, I absolutely know and understand the arguments against using AI. The work I'm doing is to mitigate the side-effects of an aggressive and haphazard rollout of AI capabilities into federal agencies, which is already happening. This is harm-reduction work that I have an opportunity do in my position near the intersection of AI and the federal government, using the skills and latitude available to me, to influence how others approach their situation. Would you rather no one did this work?
It low-key surprises me how unity never learns. The engine is so damm good but they're managing it so poorly
AI may be more useful when it challenges our thinking instead of simply reinforcing it. @TheConversationUS looks at how using AI to surface opposing perspectives and question assumptions can support critical thinking and lead to stronger ideas. twp.ai/4hwr6P
Can anyone explain what Bengio means when he demands "highly capable AI that is trustworthy by developing systems without any preferences and related objectives." www.ft.com/content/7afa... I don't get what he's getting at
