Let's say I have about 100k novel-length texts I want to import into a model like this (I want to build a book recommendation system). How should I approach it? I think this product can't deal with that much text yet?
Show HN: Customizable, embeddable Chat GPT based on your own documents
111–120 of 135 posts
Re: Show HN: Customizable, embeddable Chat GPT based on your own documents
#112Congrats on the launch. We are building something similar at https://trypromptly.com that allows chatbot building on user's own documents powered by GPT. We also allow users to build AI powered apps without writing code. For example, https://trypromptly.com/app/9da10a4f-6d20-431e-98f7-048fab81... is a chatbot built on Coursera video transcript. Similarly https://trypromptly.com/app/d478594d-2082-46c9-bee7-f057f4bc...…
I feel it is impolite to hijack someone else’s Show HN to promote your own, like-for-like product, that too as a top level comment. Not sure about HN norms around this though.
Re: Show HN: Customizable, embeddable Chat GPT based on your own documents
#113I broke it trying to upload a PDF but that is ok, I'll try it again at some point.
Re: Show HN: Customizable, embeddable Chat GPT based on your own documents
#114R.I.P. software, now there are only services (backed by software) but the software is not longer distributed, just access to the service enabled.
That being said, it seems like an almost inescapable consequence of the fact that the software itself is also data and as such subject to a similar problem: in times of ubiquitous high-speed internet access where you can't protect software (data) from being widely copied and shared over the network once anyone else gets hold of it, the only simple and reliable way for a software company to keep control over the distribution of their software (and thus its monetization) is to never share it in the first place but only provide access to it as a service.
So in the end, the shift from locally installed software to remote services kinda comes down to the fact that most users don't have the same level of awareness or care about their own data and keeping control over it as the software/internet service companies have for their data (e.g. software).
Re: Show HN: Customizable, embeddable Chat GPT based on your own documents
#115Re: Show HN: Customizable, embeddable Chat GPT based on your own documents
#116Re: Show HN: Customizable, embeddable Chat GPT based on your own documents
#117R.I.P. software, now there are only services (backed by software) but the software is not longer distributed, just access to the service enabled.
Indeed, the ongoing fashion shift from locally installed software to external services is quite problematic, as it shifts the control over the user's data from the user themself to… whatever tree/graph of third parties somewhere in the network where the data is accessed, processed… and further shared (willingly or not, knowingly or not, in conformance with the agreements and laws or not) which the user has no control…
Re: Show HN: Customizable, embeddable Chat GPT based on your own documents
#118Earlier quoted context omitted.
> Do you expect every webapps you use to tell you if they use Azure, AWS or GCP? In their privacy policies, yes. And I expect them to have signed a DPA with those and other vendors.
I’m downvoted but I still believe the position doesn’t make sense. You state and negotiate the level of protection you want for your data with your direct interlocutor. They then enforce that with their own suppliers. You are free to want to know the whole chain of companies involved in providing you a service but that’s very unusual.
https://cloud.google.com/terms/subprocessors
https://aws.amazon.com/compliance/sub-processors/
https://www.cloudflare.com/gdpr/subprocessors/
https://www.atlassian.com/legal/sub-processors
https://slack.com/slack-subprocessors