AI Project building software. Rather than make a 18 month schedule with p6 or Project, have the AI go through a database of all past construction projects and build a skeleton based off of past work automatically with the ability to adjust automatically for changes in the scope of work with human oversight and finalization.
Make 18-24 month projects in P6 or Project is quite the task. If I could have an AI software go through the companies past projects and then automatically fill in a baseline schedule based off of past work done. Also be able to automatically adjust for changes in scope of work or delays. Obviously any intense schedule is going to require human oversight but laying the base skeleton down automatically and adjusting some settings would be amazing.
A lazy way to solve the fact I drown in my own photos/videos/screenshots/etc on a daily basis and don't really have time to sort through it all.
All it would need to do is group images. For example, tell concert imagery apart from pictures of friends or scenery. It could infer a lot just from the dates on the imagery, and group accordingly.
Bonus points if it can then sort through a particular group and suggest the highest quality images of the bunch based on things like focus and exposure.
Saw this project the other day when looking through go resources, sounds like it's at least part-way to what you describe: https://github.com/photoprism/photoprism
Disagree, there's something categorically different about 'AI' - technology that can analyze and generate digital data from dense data - and a programming language.
They're obviously different things, but that's not the point I was making, now was it?
The point is, what actually delivers value is that the core functionality of whatever you're building can be smarter. The fact that it's driven by AI is irrelevant to the user -- as is the programming language that you select to build the thing, the GUI library you use, etc.
People want intelligent software - the end user often doesn't care if it's AI or a proprietary algorithm - it's technomagical and as long as they have a button to press that get's them 90% of the way to their goal, they'll buy it.
AI classification and prediction of social relationships, perhaps.
Something that could read the mass of communication data in my life; social networks, email, texts. Then process it and make predictions and discover patterns over time... even determine when I was happy, sad, or angry and how the people around me felt...
(giving all your personal data to an AI is definitely not suitable for everyone)
11 comments
[ 9.5 ms ] story [ 38.6 ms ] threadWork for a GC
Make 18-24 month projects in P6 or Project is quite the task. If I could have an AI software go through the companies past projects and then automatically fill in a baseline schedule based off of past work done. Also be able to automatically adjust for changes in scope of work or delays. Obviously any intense schedule is going to require human oversight but laying the base skeleton down automatically and adjusting some settings would be amazing.
A lazy way to solve the fact I drown in my own photos/videos/screenshots/etc on a daily basis and don't really have time to sort through it all.
All it would need to do is group images. For example, tell concert imagery apart from pictures of friends or scenery. It could infer a lot just from the dates on the imagery, and group accordingly.
Bonus points if it can then sort through a particular group and suggest the highest quality images of the bunch based on things like focus and exposure.
AI is an implementation detail, not a feature. Yes, it's an excellent technology that has many different applications. Yes, it can do amazing things.
The point is, what actually delivers value is that the core functionality of whatever you're building can be smarter. The fact that it's driven by AI is irrelevant to the user -- as is the programming language that you select to build the thing, the GUI library you use, etc.
3d animation generation (walks etc) or auto poly-reduction, or auto-rigging. Especially if it could auto-generate those from videos/pictures
Something that could read the mass of communication data in my life; social networks, email, texts. Then process it and make predictions and discover patterns over time... even determine when I was happy, sad, or angry and how the people around me felt...
(giving all your personal data to an AI is definitely not suitable for everyone)