sebastiennight

51 ms·

1. 1141.▲
   by sebastiennight 4y ago

   Ah, it looks great. It's just 100% English though, which doesn't fit my use case (customers in 6 languages)
2. 1142.▲
   by sebastiennight 4y ago

   What do you use for TTS? I'm designing an in-app chatbot and having a hard time figuring out a fast-enough TTS library. Currently considering either pre-recording all the outputs (ugh) or using the browser's TTS api (ugh^2) :)
3. 1143.▲
   Ask HN: A formula or library to generate color palettes?

   2 points by sebastiennight 4y ago|2 comments
4. 1144.▲
   by sebastiennight 4y ago

   Thanks for your reply! A. OK, I like that, thanks B. OK C. Just to confirm - the chunks are verbatim right? So in theory I could just do a string search in the document to locate the chunk? D. I would assume that you are currently not encry
5. 1145.▲
   by sebastiennight 4y ago

   First: Congratulations! I've just sent an early access request. Second - I've got questions! Here is my use case. We have hundreds of clients who each have dozens of videos on our platform. Videos are grouped in collections. For e
6. 1146.▲
   by sebastiennight 4y ago

   Aha, I've also resorted to replacing the syllable "smart" with "spy" in the names of devices (since the difference between a "normal/dumb" device and a "smart" device is always that the &quo
7. 1147.▲
   by sebastiennight 4y ago

   That is a smart troll website and SEO/marketing strategy. I think the underlying challenge is not going to be easily solved. If the ChatGPT API is indeed more powerful than `text-davinci-003`, then the predatory pricing creates a ton o
8. 1148.▲
   by sebastiennight 4y ago

   Thanks for this perspective: 6. "Y" is a pretty big deal because it's a feature that makes the service works across borders: without it, the enterprise company can't onboard their entire international workforce. However
9. 1149.▲
   by sebastiennight 4y ago

   (I have to add that they said they'd be willing to fund a significant sum of their fundraise towards development of the feature Y if we partner with them)
10. 1150.▲
    Ask HN: B2B firm needs our tech to raise their $5M Series A

    5 points by sebastiennight 4y ago|3 comments
11. 1151.▲
    by sebastiennight 4y ago

    I think Noah Kagan's blog is literally called "OK Dork", and I always assumed the title was self-deprecating (in a fun/positive way) rather than negative.
12. 1152.▲
    by sebastiennight 4y ago

    LOL... I just tried it with the synopsis of the story I just wrote with GPT[1] : > When Sébastien and David discover the seemingly-magical powers of the concierge service offered by David's Mastercard, they are amazed. However, thei
13. 1153.▲
    by sebastiennight 4y ago

    Wow, this is freakishly smart. I just posted about how I figured out a way to get GPT to write a multiple-chapter novella, but I didn't know you could do something similar with images and have some stability/continuity in the look
14. 1154.▲
    Show HN: I wrote a novella with GPT-3 (with prompts and screenshots)(indiehackers.com)

    2 points by sebastiennight 4y ago|0 comments
15. 1155.▲
    by sebastiennight 4y ago

    It's very interesting, thanks! I've noticed (on the Spock example) that "make him smile" didn't produce a very... "comely" result (he basically becomes a vampire). I was thinking of deploying something lik
16. 1156.▲
    by sebastiennight 4y ago

    Wow that was a fascinating read! Thank you. It doesn't explain why the model only refers to the documentation though. Basically what they are doing is giving GPT-3 a prompt that includes the (semantically relevant) pieces of the docume
17. 1157.▲
    by sebastiennight 4y ago

    I'm very curious as to how they are able to limit the chat bot's knowledge to their own documentation. It seems as if it should also have the entire knowledge of the LLM, but > Who was Thomas Jefferson? Outputs > Hmm, I&#x27