Emily Hunt

~50% of people (in the US) use LLMs regularly, and this has only been increasing more over time. These products are already somewhat ingrained in society, and informing people about the risks and issues with tools that they are attached to does not work if you scream at them.

Americans and AI 2026: Chatbots, Smart Devices and Views on Impact pewresearch.org

if you are a man who builds apps then i have found the star for you it's called UX Men (short for UX Mensa) and you are a detached eclipsing binary star

UX Men: Detached Eclipsing Binary as a Benchmark Candidate arxiv.org

Surely this example just illustrates my point, though? Petrol, as a technology, still exists today; just it was regulated to not be allowed to contain lead. *Half* of US adults are using chatbots per week, and ~25% daily. Hence, I think calling for AI use to stop altogether is…

About a quarter of U.S. adults use chatbots daily pewresearch.org

I find it particularly shocking to see so many people so in favour of AI as a research tool - especially Claude code - even while knowing that Anthropic are sending money towards SpaceX, a company that has made it a mission to destroy the night sky for ground-based astronomy. …

Anthropic and SpaceX announce major partnership as AI arms race continues nbcnews.com

I am SO sick of headlines like this. The public sees "AI" as being LLMs/ChatGPT, not an ML algorithm - and articles like this mislead the public into thinking that LLMs are discovering novel things. It's a cool thing! I'm excited! But it is lazy reporting to suddenly use "AI" as such a broad term

I am determined to prove that you can be largely anti-AI but also not be an asshole. The attitudes of some people on this site will NOT win over any minds or win you over any favours. If you insult someone, you are doing it to a real person. And it sucks!

hot take: I would rather read your imperfect English in your own voice than perfect LLM English that makes all writing narratively the same

I honestly hate watching talks made with AI. I hate it when speakers brag about how much time it saved them. If you have 80 people in your audience and they all watch you for 30 minutes, that's 40 hours of person-time. And you're telling us you're *proud* of how little you prepared? 🤷‍♀️

Testing it out of curiosity, I feel like Google's new AI search mode sides with whichever papers make the strongest claims (regardless of whether or not they're true.) This has major, shit implications for information retrieval - the most hyped claims are rewarded with AI search rating you higher

it would be so much easier to be excited about the immense technological progress that LLMs bring to the table if the frontier labs weren't total assholes

Prices on new servers at Hetzner have increased 3×, which is pretty insane. I kinda get it because they're a budget provider that already had a tight margin - but jeez, I am really done with how much AI has inflated hardware and compute costs

Somehow, the pro-AI crowd is convinced that almost everyone is already using AI and loves it, and the anti-AI crowd is convinced that almost everyone is rejecting AI and hates it Both scenarios can't be true. It's unpopular, dangerous technology, but I think we're already stuck with it

Hot take: right or left, both sides of the political spectrum are vulnerable to disinformation, and we see that in full swing with the attitudes towards AI that some people on the left have

I sometimes wonder if it's pointless to keep holding out on AI coding, but my god I just cannot bring myself to send money to US tech companies right now