MrCoffee7

I don't understand how your editor is supposed to work and it does not have online help? For example, how can I place my cursor at a particular place in the document and begin editing the text around my cursor location?

What assumptions do the articles make about the previous level of understanding of physics that the reader has? Also, I'm a bit lost and find the articles not so "clear" when I just jump in to read a random one - is there a suggested learning path or sequence to read the articles in where they sequentially build up knowledge?

Google's AI search overview is designed to quickly pull and summarize information from its massive web index, while ChatGPT search focuses on providing detailed conversational responses that may require more processing time. The speed difference users notice comes from fundamental differences in how these systems work - Google leverages its existing search infrastructure and pre-indexed web content, while ChatGPT processes queries through a more complex language model that generates responses token by token. Also, I would imagine that ChatGPT is using RAG more in generating some of its responses, and RAG is I/O bound. I/O bottlenecks are orders of magnitude slower than a process that could be completed mostly in memory.

This sounds like doomscrolling on steroids.

A lot of the ratings numbers seem so close together that I'm not sure whether the difference between them is semantically and statistically significant?

I doubt that would happen in the U.S. - it seems like there would be too many liability issues around the possibility of innocent bystanders getting injured by a drone in pursuit of someone. Also, a lot of employees are getting crap wages and would not want to take the risk of injuring someone if they were personally piloting the drone because they could be personally sued for injuring a bystander in the drone pursuit. If fact, the store employee in this case could even be sued even if they weren't piloting the drone themselves, eg. if they were the one who only reported the shoplifting.

How does Windows ML compare to just using something like Ollama plus an LLM that you download to your device (which seems like it would be much simpler)? What are the privacy implications of using Windows ML with respect to how much of your data it is sending back to Microsoft?

Books aren't really going to help you lose weight but joining a program will. I recommend TOPS (Take Off Pounds Sensibly) which has many local chapters across the U.S. and costs much less than Weight Watchers.

What algorithms does Wiserank use to rank creativity in papers?

I think there is a strong possibility that the worsened geolocation abilities will continue as there is a tradeoff in AI models between accuracy and AI safety/risk concerns. AI companies will be under increasing pressures to boost AI safety which may well cause a hit to AI accuracy.

Does the college you just graduated from have a job placement office that could offer you help or advice? Some local public libraries have short courses where they give you feedback on your resume, etc. You could also look at libguides on job finding resources such as https://libguides.ccsu.edu/career/jobhunting . A "libguide" is a "library guide" that is a list of resources pertaining to a particular subject. Also, does your college have an alumni association that you could join and network with?

https://buttondown.com/ainews/rss is a better summary.

Your extension would not install on my Windows 10 machine. I have Chrome Version 140.0.7339.81 (Official Build) (64-bit). There was no error message - the install process just "hung" without ever completing.

This looks like a great book but it does not discuss double machine learning and does not have an index?

https://techurls.com https://skimfeed.com

Reddit has quite a few discussions on this topic that are pretty informative.

What added value would your app have beyond what I could find for free using the business databases that public libraries subscribe to like Hoover's, etc.?

Those are some pretty sweeping allegations made in the 18 bullet points. It would be interesting if you could run an AI summarizer to list which points were raised during Lex's podcast that support each of the 18 bullet points, and then run some kind of AI fact checker on each supporting argument. That makes me wonder if there are any people or discussion forums that are already doing fact checking of Lex's podcasts? It looks like some researchers are beginning to study the issue of fact checking podcasts, eg. https://arxiv.org/abs/2502.01402 .

This game really looks fantastic! Thanks for posting about it.

What are UARS and RERA?

It hallucinated when I asked about the latest recommendations for type 2 diabetes treatment from the Amer. Diabetes Assn. 2025 conference, and falsely said that the conference had not taken place yet.

Have you compared your system to https://pubmed.ai on some of your sample questions? That would give you a benchmark and perhaps give you some ideas for your application?

Is this app targeted towards clinicians or patients, or both? If both, I would suggest adding an option for whether you wanted an answer targeted towards the clinician or patient to account for differences in understanding of medical jargon, etc.?

GLM-4.5 from https://chat.z.ai because it renders math notation so well that other LLMs tend to mangle and it uses color coding when showing Python code.

There is a simpler explanation of the cited article here "Los Alamos team finds a new path toward quantum machine learning" https://www.lanl.gov/media/news/0724-quantum-machine-learnin...

Some of the stuff at the bleeding edge though like gene therapy is so expensive at millions of dollars per patient that health insurance companies are not covering it, so that stuff is not actually worthwhile to that many patients. The vaccine tech is covered by health insurance.

The A-Z sort option does not seem to work at all?

This awesome-math Github repo has a large list of many learning resources https://github.com/rossant/awesome-math

Non-paywalled article on this case: https://fortune.com/2025/07/24/north-korean-it-workers-chapm...

Your tool looks like a great start, but I think it could use more types of search filters. A vibe type filter for the kind of tone the blog has would be useful, eg. educational, entertainment, provoking, relaxing, etc. A screenshot of part of their home screen shown on the info card for the blog would also help show the visual asthetic and vibe of the blog. It would also be nice to have an exclusion filter if you don't want to see certain subjects or people. Depending on the personal and compute time that you have for this, you could even get fancier and have filters for the average estimated reading time of the blog (eg. do you want a blog "snack" or a blog "meal") and its reading difficulty level?

https://www.pubmed.ai will generate a deep research report on that specific query involving checkpoint inhibitors and non small cell lung cancer for free. It won't have the nice comparison table that Elicit can generate. Also, pubmed.ai cannot parse a query as complex as the one used for Elicit, but you can paste the Elicit query into OpenAI's chatGPT and ask it to translate the query into a form that Pubmed can parse. Pubmed.ai was able to parse the chatGPT generated query string "("Checkpoint inhibitors"[Title/Abstract] OR "immune checkpoint blockade"[Title/Abstract])AND ("non-small cell lung cancer"[Title/Abstract] OR NSCLC[Title/Abstract])AND ("platinum-based chemotherapy"[Title/Abstract] OR cisplatin[Title/Abstract] OR carboplatin[Title/Abstract])AND (bevacizumab[Title/Abstract] OR "anti-VEGF"[Title/Abstract])AND ("PD-L1 expression"[Title/Abstract] OR "PD-L1 level"[Title/Abstract])AND (effectiveness[Title/Abstract] OR efficacy[Title/Abstract])AND (safety[Title/Abstract] OR toxicity[Title/Abstract])."