Some communities are particular about maintaining cleanroom reverse-engineering (case in point: Portal 64 getting taken down for using Nintendo's leaked N64 SDK and not the OSS libdragon SDK) and are wary of people submitting AI-generated PRs without being able to account for their sources
I think it is worth asking whether anthropomorphization of LLMs does anything to contribute to a better understanding of them that isn't entirely post-hoc ("of course they were going to find CVEs, solve ARC-AGI...") or if it just ends up reinforcing simplified teleological views of neuroscience, evolution or psychology etc.
Sure, there are potential risks to anthropomorphizing language. There's also potential benefits, in terms of allowing us to talk about an emerging alien intelligence.
FWIIW, I don't think I was anthropomorphizing. Stretching an analogy, yes. Optimization pressure is producing something with at least characteristics of intelligence. Obviously in a very different way than evolution produced the human brain, but the similarities are there.
Presumably that HN has an automated LLM-based filter to moderate rude or irrelevant comments, which is a little irrelevant to the article's wish not to see generated posts or comments, and not without its own problems besides: ML is necessary for large-scale moderation, but there are guidelines (like "don't post AI-generated text") which are IMO necessary but also just definitionally hard to enforce in an automated fashion
Yes, even though fast food places would previously use every food photography trick in the book to make the food look good, there was at least some relation between the menu and what you actually get, which people just seem completely fine with dispensing because the affordances of image generators work against including product photography
(Have also heard about people - not just scammers/spam operations - using generated pictures and descriptions for dating profiles and real estate ads, and I cannot understand what outcome they expect when someone follows up and immediately realizes what's up)
> - how many people in a team do you need and how long would it take to write a browser engine from scratch
The boring not-really-an-answer is "it depends on which sites you want to render properly": a browser capable of rendering Hacker News is a perfectly fine one-person project (which the book would get you to), supporting things like banking and online mail requires not just developers but also people willing to "dogfood" test it as their daily browser
I can't really see a definition equating the two that realistically accounts for multiple models from different vendors - am I at fault if a particularly poorly implemented harness misinterprets me writing "do not run dd if=/dev/zero of=/dev/sda" as a command to run?
In the case of TCRF, agents were also actively ignoring robots.txt and disregarding the site's instructions for interacting with it appropriately (with the "delete all data" instruction being specifically hidden from the discoverable page) - so I can't really see the case for comparing it to malware that deliberately tricks users into running it
AI does make cybersecurity more difficult, and that also goes for people using coding harnesses, sandbox your projects!
I think the difficulty was specifically with CAPTCHA challenges, which had to be quick to generate but still legible - OCR on physical documents has to be robust to a different set of problems
That said, what kind of errors are you getting? I think the main practical difference between Tesseract's LSTM-based OCR and newer VLM-based OCR like PaddleOCR [1] is (hopefully) getting to skip making heuristics for the layout of the document, but errors possibly compounding over multiple tokens - are you getting individual illegible words or having the documents smooshed together because the OCR can't parse the layout?
>> no matter how I scan receipts or documents, the OCR output seems far worse than human
> what kind of errors are you getting?
Just noticeably worse character recognition, particularly where the document is faded, water-stained, or the paper document (not the scan) was low resolution to begin with, as compared to 'normal' human recognition.
I was largely using Tesseract in conjunction with the self-hosted Paperless-NGX, and I wanted to stay free/local without yet investing in AI-focused hardware. But you're right that AI will continue to advance, including in smaller local models, and it looks like Paperless 3.0 released recently (after my testing earlier in 2026), including with AI functionality.
> or having the documents smooshed together because the OCR can't parse the layout
I haven't even been worrying about that yet - I'm just at the point of trying to get the OCR characters right. :-)
> We could train a model on a dataset where any reference to consciousness is thoroughly removed. Then explore whether the model can bootstrap a discussion about consciousness.
I'm less interested in the philosophy than the practical "how directly does training data influence model behavior" question, however PleIAs' SYNTH dataset [1] used to train the Baguettotron model has a small set of "self-awareness about training condition" documents which are amplified to (attempt to) give the expected answers to "what are you?" questions, filtering out any rows with "Pleias self-knowledge" under "query_seed_url" might be a good first stab at such a dataset
Yeah, as someone used to searching for specific turns of phrase, having the query automatically interpreted as a question is quite annoying - for looking up song lyrics, game walkthroughs and articles by quotes I want a page list, not a summary
There's a RAM crisis on, you know, let me save you the compute!
reply