droidjj

8 ms·

151.
Claude Visual and Interactive Content(support.claude.com)
2 points by droidjj 7mo ago|0 comments
152.
by droidjj 7mo ago
I recently turned my unused Google Pixel 8 into a server for my personal site and various side projects. It's super satisfying to spin things up in a couple hours, point a cloudflare tunnel at it, and share it with the world.
153.
Enhanced Atkinson Hyperlegible Font (2025)(brailleinstitute.org)
2 points by droidjj 7mo ago|0 comments
154.
by droidjj 7mo ago
Huh, it's working for me (on Firefox).
155.
by droidjj 7mo ago
The plaintiff is represented by one of the largest law firms in the world, Sidley Austin. Which is to say: they have legal firepower.
156.
OpenAI hit with lawsuit claiming ChatGPT acted as an unlicensed lawyer(reuters.com)
16 points by droidjj 7mo ago|6 comments
157.
by droidjj 7mo ago
> "Writers win the Pulitzer Prize and sell just [a] few hundred copies." For anyone else who was intrigued by this statement: The essay links to another Medium essay[0] which links to a book critic's blog[1] which links to
158.
Ungrounded Divergence–Philosophical Framework for Understanding AI Hallucination(download.ssrn.com)
3 points by droidjj 7mo ago|0 comments
159.
How to Count AIs: Individuation and Liability for AI Agents(papers.ssrn.com)
3 points by droidjj 7mo ago|0 comments
160.
by droidjj 8mo ago
*without?
161.
by droidjj 8mo ago
I’m all for regulation of AI, but that’s not a serious solution where the problem is the government pressuring private companies to do evil things. Consumer pressure isn’t much, but it’s not nothing.
162.
When Chatbots Are Used to Plan Violence, Is There a Duty to Warn?(nytimes.com)
6 points by droidjj 8mo ago|0 comments
163.
by droidjj 8mo ago
It's a nice reminder of how impressionable kids are. A little encouragement can go a long, long way.
164.
AI chatbots provide less-accurate information to vulnerable users(news.mit.edu)
3 points by droidjj 8mo ago|0 comments
165.
by droidjj 8mo ago
> We find the LLM to be perfectly formalistic, applying the legally correct outcome in 100% of cases; this was significantly higher than judges, who followed the law a mere 52% of the time.
166.
by droidjj 8mo ago
Whether it’s reassuring depends on your judicial philosophy, which is partly why this is so interesting.
167.
GPT-5 outperforms federal judges in legal reasoning experiment(papers.ssrn.com)
310 points by droidjj 8mo ago|238 comments
168.
by droidjj 8mo ago
I have been looking for an OCR model that can accurately handle footnotes. It’s essential for processing legal texts in particular, which often have footnotes that break across pages. Sadly I’ve yet to encounter a good solution.
169.
by droidjj 1y ago
The author’s main point is that there hasn’t been an uptick in total code shipped, as you would expect if people are 10x-ing their productivity. Whether folks admit to using AI in their workflow is irrelevant.