A powerful AI to-do list app needs to be built.
Hustle Guy
24 replies
• Figures out the priority of your tasks, surface things at the right time
• Suggests and creates plans to tackle tasks
• Regularly self-cleans by suggesting old tasks to delete or archive
• Auto suggests grouping related tasks
Replies
Valorie Jones@val_jones
Agreed. Right now, my to-do list lives in Monday.com - but sometimes it can't compete with a good old-fashion piece of paper that give you the satisfaction of crossing a finished task. What I need is a light weight, intelligent tool that links to all my project management software.
Share
Hey, I'm actually building a Todo App with AI at the core. The features you have suggested are very interesting.
I'm planning a beta release within a few weeks.
agreed, I believe that creating a highly effective AI-powered to-do list application is a challenging yet worthwhile endeavor that could greatly enhance people's productivity and organization.
I completely agree! A smart to-do list app that adapts to my behavior is a game-changer. It would save me time and help me focus on what's truly important. Can't wait to see this built!
That would be a killer idea! Are you working on something like it? :)
It's impossible to do it, you will have to constantly feed it with information :( however, you can definitely use AI to help you identify tasks that you might overlook. But again this is not a streamlined process and many times it's a pain to do it. I used chatGPT4 for this, it does the job but it doesn't cover all the details.
@laurentiu_cotet you can constantly feed it with information. If you can think it, you can say it and if you can say it speech-to-text can transcribe it and make it accessible to the AI. Your note was my initial reaction too. But it's very doable.
@stephen_smith67 That would be nice, but what we think about a certain task and how we feel about it might vary depending on the day, our mood, and other factors. Therefore, the information we feed to AI might not always be accurate. I have had experience with various mood trackers over the course of a few years, and I found that they were not accurate at all.
@laurentiu_cotet thanks for the thoughtful response but can I suggest that the to do list should be just as dynamic as your mood - and also remind you that inspire of your mood certain things that you identified that need to get done, still need to get done. Part of that can he helping to defer the task until you few more like it. Rather than lose it altogether. But you are right mood has so much to do with our productivity .
I guess the only way this works is using a NLU model that can listen to priorities and tasks, organize them, provide a timetable. More importantly it could do a bunch of the tasks which would be more than helpful. So 'write to George Smith by EOD about the new contract', gets drafted and put into email drafts and a prompt at 4.30Pm to check and send! I am beginning to like this! Go fo it someone!
I'd imagine something like this could be integrated into existing note taking tools like Notion
Would be interesting!
I agree! I think we'll definitely start seeing a lot more productivity apps integrating AI into their user flows.
Agreed. That’s what I’m trying to build with Heyroger.xyz. Still early days and I’m hoping to get some feedback
I think this could be a very useful idea. However, for it to actually cater anything to you specifically, you would have to allow the AI to have access to virtually every aspect of your life. From when and how you wake up, to your mood during the day, to your location and your interactions with others. Any thoughts on that?
NLP/NLU transformer models are becoming smarter every day. Recently, some techniques have been developed to improve their performance even further. For example, one technique called "knowledge distillation" helps to transfer information from larger models to smaller models, allowing them to perform better. Another technique is "pre-training," which helps models learn from a lot of data before being fine-tuned for specific tasks. There are also new ways of representing text called "contextualized embeddings" that can help models better understand the meaning of words in different contexts. All of these techniques are making NLP/NLU transformer models better at understanding and processing human language.
You can follow a good source/ai developer blog for more information: https://www.soffos.ai/developers...