How, What, Why, and Which of Real-Time Data in Msty

In version 1.0 of Msty, we introduced a new feature called Real-time Data. This feature allows you to get real-time information from the web, so you can get the most up-to-date information on any topic. This is very similar to Perplexity (if you know, you know), but with the power of using any model you want, a great UX experience, and keeping your data private.

How does it work?

The way this feature works is quite simple. When you send a message with Real-Time Data enabled, Msty fetches the most recent information from one of the popular search engines, selects the top few results, cleans it up, and feeds it to the model. The model then generates a response based on this information and provides you with the AI's response along with the sources it used. That's it! No magic, just a simple fetch, clean, and feed process.

Of course, there are some optimizations and tweaking involved to make this process faster and more efficient. And there are fallbacks in case one search engine is down, or it is blocking you from fetching results, or the response is not as expected, or empty.

Msty goes through a lot of effort to ensure that your request gets blended in to make it look like it's coming from a real user and not a "bot". This is to make sure that the search engine doesn't block you from fetching the results. Making this blend-in work took us more time than actually implementing the feature! Still, we recommend that you use caution and use this feature sparingly, and not misuse it. We don't want you to get blocked by the search engines.

What about privacy?

We understand that privacy is a big concern for many users, and we take it very seriously. That's why we have ensured that the data fetched from the web is not stored anywhere and is only used to generate the response. We don't store or track any of the data that is fetched from the web. We also don't share any of this data with third parties. It doesn't go through any middleman servers or anything. We send your query, get results, send the combined data to the LLM that you have selected, and show you the response. That's it. We have no further business until you send another query.

However, your query does get sent to search engines, and they might store it. We have no control over that. This is no different from you searching for something on the web yourself but, arguably, more private because there is no tracking, no analytics, no third-party scripts, no cookies, etc.

Why not do X to protect privacy?

Real-time data means we have to fetch the results from the web in real-time. This means we have to send your query to the search engine to get the results. So unless we are going to build our own search engine, we have to send your query to the search engine.

Even if we were to scrape results from some web pages, we need to know how to get to those pages. Someone suggested that we scrape the URLs directly. Sure, but how should we be getting those URLs? We need a central index that we could query for the relevant results, and there are billions of them. That's exactly what a search engine does. So, we are back to square one! We have to send your query to the search engine. This is the most efficient and accurate way to get the results. And we have to do this in real-time.

Yes, there are other ways such as having a local index, training your own model periodically, etc., but none of those are feasible and within the scope of Msty. That is going to take much more effort and going to distract us from our main goal and implementing other important features. Msty is not a search engine; it's a conversational AI app. We want to keep it that way.

Very soon, Msty will warn you about your query being sent to search engines when you enable it for the first time. As much as we hate to add popups and dialogs (ugh), we think this is important for you to know.

Which search engines do you use?

We use multiple search engines to get the results, and we rotate between them to make sure that we are getting the best results. We also have fallbacks as mentioned above. Right now we are using Google and Brave Search, but we are planning to add more in the future. We might also consider adding a setting where you can select which search engine you want to use. But, having settings for everything is not always a good idea and goes against our principle of things should just work without having to configuring too many things. We want to keep Msty simple and easy to use, so we will only add settings if it is absolutely necessary and there is a high demand.

We are always in the lookout for ways to improve this feature and make it more useful for our users. If you have any suggestions or feedback on how we can improve this feature, please let us know. We are always open to feedback and suggestions from our users, and we want to make sure that we are providing the best possible experience for our users.

If you are still concerned about privacy, you can always disable this feature and use Msty without it. There are plenty of other reasons to use Msty other than Real-time Data, if not we've failed in our mission. We want to make sure that you are comfortable using Msty and that you have the best possible experience with our product. If you have any questions or concerns about this feature, please let us know. We are always here to help, and we want to make sure that you are happy and feel safe and protected using it. Thank you for using Msty!