Strange, I am running it on a Snapdragon 8 Gen 2 (Z Fold 5), and it's totally fine for me. (If anything, it's a little too good at staying in the background; if you have private tabs open it insists on persisting in memory.)
Not saying your issues aren't real, but rather maybe there's another app or your manufacturer's flavor of Android that's causing the issue (like those aggressive background killers).
As for Edge, I used to be a big fan, but when they finally introduced history and tab syncing in 2021, it didn't have E2EE, and it still doesn't, which I find inexcusable. All the other major browser vendors offer it, even Google (though you have to opt in).
Interesting project. I've been thinking about a tool like this; I might be following a multi-volume book series, but it's been years since the last book. When I pick up the latest volume, sometimes there are details that I just can't remember (small details that may turn out important, relationships between minor characters, etc.)
I would just consult a fan wiki, but that doesn't work if the title isn't popular or if the book is too new. This seems like the perfect tool if it can somehow maintain coherency across multiple books.
That said, I do understand (and share) a lot of the frustration and hesitancy that people here have around AI tools; I don't want an app that takes away the act of thinking (like that post recently about teachers using AI to make banal lesson plans, and students in turn using AI to write essays -- what is the point then?). I hope you don't take it too much to heart, and try to showcase use cases where your app can actually provide value.
Another piece of feedback is it would be great if this could be all packaged up into a docker image that would make it easy to deploy on a local machine (or like on a home server/NAS). Right now it seems there are still a lot of manual steps and scaffolding.
> That said, I do understand (and share) a lot of the frustration and hesitancy that people here have around AI tools
I share some of the same feelings as well.
As for use cases where it can provide value, I think it can be of value if you want to read difficult academic, technical or business books with deep understanding. I think so.
> Right now it seems there are still a lot of manual steps and scaffolding.
I think you are right.
I originally planned to use it as a tool for my own exclusive use, so I was able to build an environment with minimal implementation costs, but I didn't expect to get so many comments.
I will improve it!
Why would you need to retrain the model or update the SFT? You could just dynamically update the system prompt to include things it should advertise.
You could even have something like an MCP to which the LLM could pass "topics", and then it would return products/opinions which it should "subtly" integrate into its response.
The MCP could even be system-level/"invisible" (e.g. the user doesn't see the tool use for the ad server in the web UI for ChatGPT/Claude/Gemini.)
> Used as supplied, Google Tag Manager can be blocked by third-party content-blocker extensions. uBlock Origin blocks GTM by default, and some browsers with native content-blocking based on uBO - such as Brave - will block it too.
> Some preds, however, full-on will not take no for an answer, and they use a workaround to circumvent these blocking mechanisms. What they do is transfer Google Tag Manager and its connected analytics to the server side of the Web connection. This trick turns a third-party resource into a first-party resource. Tag Manager itself becomes unblockable. But running GTM on the server does not lay the site admin a golden egg...
By serving the Google Analytics JS from the site's own domain, this makes it harder to block using only DNS. (e.g. Pi-Hole, hosts file, etc.)
One might think "yeah but the google js still has to talk to google domains", but apparently, Google lets you do "server-side" tagging now (e.g. running a google tag manager docker container). This means more (sub)domains to track and block. That said, how many site operators choose to go this far, I don't know.
Slightly related I've also been recently noticing some sites loading ads pseudo-dynamically from "content-loader" subdomains usually used to serve images. It's obnoxious because blocking that subdomain at the DNS level usually breaks the site.
My current strategy is to fully block the domain if that's the sort of tactic they're willing to use.
There are commercial services that provide residential proxies, i.e. you get to tunnel your scraper or bot traffic through actual residential connections. (see: Bright Data, oxylabs, etc.)
They accomplish this by providing home users with some app that promises to pay them money for use of their connection. (see: HoneyGain, peer2profit, etc.)
Interestingly, the companies selling the tunnel service to companies and the ones paying home users to run an app are sometimes different, or at least they use different brands to cater to the two sides of the market. It also wouldn't surprise me if they sold capacity to each other.
I suspect some of these LLM companies (or the ones they outsource to capture data) some of their traffic through these residential proxy services. It's funny because some of these companies already have a foothold inside homes (Google Nest and Amazon Alexa devices, etc.) but for a number of reasons (e.g. legal) they would probably rather go through a third-party.
• Inventing on Principle (https://vimeo.com/906418692) / (https://news.ycombinator.com/item?id=3591298)
• Up and Down the Ladder of Abstraction (https://worrydream.com/LadderOfAbstraction/)
• Learnable Programming (https://worrydream.com/LearnableProgramming/) / (https://news.ycombinator.com/item?id=4577133)
• Media for Thinking the Unthinkable (https://worrydream.com/MediaForThinkingTheUnthinkable/)
Or you could just check his website: https://worrydream.com/