I tried it to summarise some of my own blogs posts, it is pretty good. Sometimes the summary makes the same point several times, and its tendency to insert content not in the source material makes it hard to trust.
Universal Summarizer
141–150 of 172 posts
Re: Universal Summarizer
#142this is really awesome. Especially for people who are dyslexic or can't concentrate for a while by reading longer texts. At least they will know what the article is about. I also like the "key points" button.
Re: Universal Summarizer
#143Tried it with code: https://bellard.org/quickjs/pi_bigdecimal.js This code uses the QuickJS bigdecimal type to calculate the value of pi to a given precision. It does this by using the Chudnovsky algorithm, which is a series of calculations that can be used to approximate pi. The code is written in Javascript and uses BigInt and BigDecimal to perform the calculations. It is interesting to note that the code also take…
no way! this can also summarize / explain code!? wow!
Re: Universal Summarizer
#144Doesn't seem to work https://labs.kagi.com/ai/sum?url=https://www.youtube.com/wat...
Re: Universal Summarizer
#145Not bad with a single out of context chapter from a genre web fiction serial that is hard to understand without context. The language and writing style sounds a LOT like ChatGPT. Feels like the model asked a bunch of different questions (what is interesting to note, what's important, etc.) >Jessica Yamada is a therapist working with a patient, Victoria, who is struggling with the possibility of uncontrollable, incura…
If it's a LLM, it's probably read a lot of webfiction.
Re: Universal Summarizer
#146This is quite a bit better than all of the other attempts at building AI summarizers I've seen. One thing that's really nice about it is that it seems to pick out surprising or unique information, rather than just trying to provide the most general abstraction or summary of the article. My team has a journal club where we go through ML papers, and I feel like this will be a good tool to familiarize myself with academ…
it's very good at that..how does it manage to do that?
Re: Universal Summarizer
#147Impressed with their long document summarization, any ideas how they do this? Seems beyond normal GPT limitations; either they have a more powerful model (doubt it) or hacked around the limitations? e.g. good summary for a very long text https://labs.kagi.com/ai/sum?url=http://localroger.com/prime...
Thanks for noticing! (dev here) We have an in-house model we've been developing since 2019, just for summarization of long documents in real time. We'll try to find some time to blog about the high level design.
Jealous. I have been trying to 'configure' my own local summarizer. As I find the ChatGPT limitations on length to be frustrating. For example, I wanted to take a transcription of a troubleshooting session with user and support agent (me). I wanted a summary of what we did essentially, overall. But doing this with ChatGPT takes so. many. copy/paste actions. I just wanted to give it all, and say, get a summary of every 5 minutes.
Secondly, this lab needs to stick around. It makes me want to subscribe, just for access to this.
Re: Universal Summarizer
#148I had it summarize a recipe for me. It did a good job, but definitely editorialized. The URL I sent included just the recipe for split pea soup, but the summary added that "Split pea soup is a hearty and comforting dish that is easy to make" (which is true, but wasn't in the original) and said that the soup "is sure to be a hit with the whole family" (again, true, but wasn't in the original). So I feel like "summary"…
This should be more visible. Both in terms of "everyone should see this comment" and "maybe Kagi should be labelling this more truthfully"
To be fair, it’s also under the `labs` subdomain, which is their experimental stuff.
Re: Universal Summarizer
#149This is a great tool for survival in the face of the popularity of independent journalism.
Perfect. Thanks!
Re: Universal Summarizer
#150I had it summarize a recipe for me. It did a good job, but definitely editorialized. The URL I sent included just the recipe for split pea soup, but the summary added that "Split pea soup is a hearty and comforting dish that is easy to make" (which is true, but wasn't in the original) and said that the soup "is sure to be a hit with the whole family" (again, true, but wasn't in the original). So I feel like "summary"…
“ Rafal Pastuszak is available throughout the week for a coffee and a chat about ideas big and small. He also has a bi-weekly 2h slot which he uses for pair programming and mentoring sessions with friends, old and new. He also offers a 3 minute rant session, where he listens to the other person rant and then they spend the remaining 7 minutes contemplating the pointlessness of their endeavors.
Rafal is interested in a wide range of topics, from product development to mental health, and from photography to emerging technologies. He is also open to pair programming with anyone on their own problem. His goal is to get to know people and what they have been up to.
It's an interesting offer for anyone looking for someone to bounce off an idea or learn something new.”
Fair enough, there’s a bit of editorialising but the main gist is fine.