To test how well a model is able to retrieve content from its long context, we use the needle-in-a-haystack test. Aside from reasoning about long documents and complex codebases, what I find to be…

Aishwary Gupta

Venture Capital

Community Picks is a section on daily.dev where our community members share the most interesting and valuable content they've discovered online. From insightful articles to handy tools, every post is a gem curated by our dedicated coomunity. To contribute to Community Picks, you need to have at least 250 reputation points, ensuring that only active and trusted members can share their finds.

Community Picks

Gemini 1.5 Pro, a large language model, demonstrates its ability to understand and reason about videos. The post discusses the needle-in-a-haystack test, a video demonstration, and the potential applications of Gemini. The author then shares their experience using Gemini to analyze a 59:59 minute video lecture, testing it on various questions. The post concludes with reflections on Gemini's summarization capabilities and provides useful links for further reading.

I used AI to watch an hour-long lecture

Why would a VC say no to your product? ❌

List down all the questions asked in this talk and what the speaker responded as a suggestion/solution ❌

finally you don’t need to spend a day to read dumb videos

So it’s still hallucinate when the questions are not specific enough, so does this all come down to prompt engineering?

There has been AI’s for some time now that summarize videos for you specially on YouTube. But I don’t find them as useful.