sorokinvj

I developed my personal site - https://vladislavsorokin.com (not much content there though) with NextJS typescript markdown blog template https://github.com/vercel/next.js/tree/canary/examples/blog-...

Now I am thinking about markdown editor, because writing content in VSCode, well, its ok, especially if you copy-paste from GPT, but ideally I would like to write it on the page itself so you see immediately what you get and you don't need to run the localhost etc. I think its not really that hard to make it.

Currently, its a Mixtral-8x7b. We are thinking about giving users an option to choose a model, but so far nobody asked us to do so :)

Look, surely it would do so, if the product become mainstream.

However, recently I suggested a Github Copilot to a dev, that was not really comfortable with filtering an array (easy task for anyone who code). This developer was on the salary in a product company (I have no idea how he achieved that). After 2 weeks I made this suggestion, he was fired. Copilot completely spoiled his undeveloped cognitive skills and he started to work even worse that he was before. Do Github Copilot team have a responsibility for what happened - I doubt it.

So yeah, tech is double-edged sword 100%. Use it wisely.

Great questions!

Let me try to answer them:

1. Intuitively, the use case "I don't have time to type in prompt, but I want to query AI real-time" does make sense, but is there a real use case where people actually would find it useful and beneficial so that they would pay for this ability - is unknown. We decided to test it in this specific niche and very narrowed down usecase.

2. Currently we use groq-sdk with Mixtral-8x7b. We tried gpt-3.5, gpt-4turbo, claude 3, but for our specific use case speed and the ease of predictable output made groq-sdk our choice for now.

3. RAG is coming, but first we would try something much simpler and launch a feature where you could upload a CV. We would parse it and add to system prompt.

4. Open source is a great idea, but I personally believe that this whole family of voice-driven, real-time ai assistants (that are definitely coming, no questions about it) need a lot of product refinement. Just to share our code now would give immediate value for developers who wants to do the same, but not for the end users. And to be honest, even our code, although being thoughtful and efficient in solving specific challenges (like async orchestration of many bg processes), I don't think it would bring a lot of value now. So yes, once the product side will have more data from users and the product itself would be bit more shaped - sure

hey! thanks for comment! could you help us investigate the issue?

but what is the real user story?

take a look at https://www.airplane.dev and https://www.scriptkit.com

Have you guys ever considered frontend app for users to actually built am organism in the browser? I have no idea if its something that anybody needs in the world but this could be fun! I would be interested building that

this is exactly how https://www.next-kraftwerke.com built one of the biggest European power generation network - they used generators from important social infrastructure points, like hospitals. Usually these generators were used only as a backup power source, so basically they've been just standing there and loosing its value with time.

The idea is definitely in the air, but I guess its hard right know to come up with practical use cases. I mean I do believe and want to control my digital trace, but at the end of the day I don't know what to do with it except maybe visualize it on virtual whiteboards like Temin - the project showcased among others several threads above.

For privacy maniacs, of course it would be awesome to add functionality and destroy your traces, but I doubt you could achieve it with current API's.

Also interesting maybe to produce high-level analysis of your personality traits based on messages-comments-articles you write (but to use how later - I am not sure)

I developed a speech recognition service for journalists couple years ago, now migrating it to voxpop.live (it was speech.media). Unfortunately, ASR I used (Amazon) wasn't good enough in Russian (my main target was Russian journalists). Now, I am thinking what to do next with it, lets get in touch and exchange ideas?

The best on the market AST in my opinion is Speechmatics, but to use it commercially you need to deposit $10k. Btw there are services that let you train your own model, i.e. https://www.defined.ai

Also as an idea, to use GPT-3 and correct any model response (punctuation, summary) but its gonna be expensive!

hey Oliver, awesome project, I will definitely try it and I am also interested in developing it further - Miro is breaking new record every month, so totally worth to compete.

Also it could be an interesting intersection between laptop (where you create data and put things to the boards) and VR (where you navigate, process and work with data). I am not sure about MVP, but surely there is so many usecases, starting from personal collection of knowledge to teamwork, project documentations etc.