101–110 of 341 posts
Looking forward to seeing GPT4-Turbo there soon.
Can anyone explain what "extra knowledge" means specifically? Is that like fine tuning? How much data can I give it to learn? How much can it retain? Can it be updated over time?
The keynote used a .txt file of a lecture that the user uploaded as a data source the model can select from. From a technical perspective, it's anoher data source for retrieval-augmented generation (RAG) doing a vector search in the background: https://platform.openai.com/docs/assistants/tools/knowledge-...
I wonder how much money I could make making "GPTs" full time. Barrier to entry is nonexistent so I imagine highest revenue ones if this becomes a serious thing people use will be advertised externally or have some proprietary info/access.
So these "GPTs" are the combination of predefined prompts and custom API access? Not customly trained LLMs? If so, I guess you can make such a "GPT" on your own server and independent from a specific LLM service by using a prompt like ...you have available an API "WEATHER_API". If you need the weather for a given city for a given day, please say WEATHER_API('berlin', '2022-11-24') and I will give you a new prompt inc…
However creating something like this previously required a Jupyter notebook but now just...asking for it. Makes it accessible to 10x more people
[dupe] More discussion over here: https://news.ycombinator.com/item?id=38166420
Earlier quoted context omitted.
I'm not sure i understand?
The GPT's making the most money will be made by larger companies who advertise use of it and maybe make it a funnel to their in-app integration, or GPTs which are made effective by information that is proprietary.
A good paper on the state of the Alexa skills BTW
SkillVet: Automated Traceability Analysis of Amazon Alexa Skills https://ieeexplore.ieee.org/document/9619970
Earlier quoted context omitted.
Imagine a "GPT" that could generate websites and provide you with a live deployment as you change it using natural language. A website builder GPT that is primed to output and design in a decent way, that has all the prep beforehand to use particular libraries, and integrations with something like Render. People would pay for that. Sh--... I better build it!
One can come up with all sorts of ideas like this, but building it will be a matter of slow iterations at prompt engineering in a mixture of natural language and data structures and will be at the whim of changing APIs, including the backing ChatGPT model. Sounds messy, hard to manage, hard to test...or am I missing what the actual process will be for creating one of these?
So i guess the next wave of startups has been killed. I'm not even sure what kind of startup still makes sense as a gpt thin wrapper?
Thin wrappers around anything never make much sense. They're trivially replaceable.