I figured out I can also just ctrl-A copy all text on a website for the A.I. to read and summarise it from the clipboard contents. That way there is no html to filter out, just titles and links, recognisable by their lack of verbs.
I suppose it wouldn’t be too difficult from that point on to have it reply to forum posts here.

Further tied the A.I. into Windows Media Player by having it search for music filenames or playlists that contain keywords from the command sentence (e.g. “Play some keyword music by keyword keyword”). Though honestly you can just as well do that with Windows Speech Macros or other virtual assistants. I also thought of having it turn the radio on and off with a WiMo power plug, but then I discovered you can just launch websites that feature live radio.

I now have an A.I. alarm clock (video). It doesn’t take any A.I. to do these tasks, but it’s easier to set by typing “wake me at 8 tomorrow” than the setup of a normal alarm clock. The weather data is downloaded and converted before they’re put in a sentence (e.g. rainfall from mm/h to a percentage of cats and dogs), and the lights are turned on by sending a command to IFTTT.com that activates a €40 D-link smartplug.

Another thing I made with help of the internet is a voice interface between my smartphone and my home PC, using the Tasker app for Android. With Tasker you can set up a sequence that launches Google Now speech recognition, sends the captured text to a php page of your own making (e.g. your chatbot’s online location), and then reads aloud the contents that the php returns, namely the A.I.‘s response. It works pretty well.

OR…you could tell Google Now, “Wake me at eight am tomorrow” and it would set an alarm for you.
(at least mine does).

I pretty much depend on Google for so many things in my life that I used to have automated via hardware and lots of electronics.

Then again, that’s just me.

I recently say a video of converting a large mirror to a daily wake-up, informational display mirror, (time, weather conditions, News feeds and many other possibilities), using a Raspberry PI or similar. You could see your face but you could also see the time, weather, etc. in the upper corners, which was really nice.

https://www.youtube.com/watch?v=fkVBAcvbrjU

I did, however like your setup.

I didn’t know Google Now could do that (I’m not prepared to open my life to Google), but I did know Siri can also. Most of what I do in terms of virtual assisting is an exercise in futility, just to see if there’s any practical use for the A.I. . A minor benefit would be that I could include dynamic conditions like “Wake me 10 minutes before the Muppet Show starts”, a feature Google Now seems to have removed. But that too is nothing a dedicated scheduling app couldn’t do.

Mark Zuckerberg also recently did some home automation and added a somewhat novel feature: By telling his program when it played the wrong music, it learned which songs matched which keywords. At least that’s what I make of his story.

I’ve come across a few smart mirrors too , they’re pretty cool but also a tad expensive to make. Here’s a simpler version made with a tablet: https://www.youtube.com/watch?v=fUcbEqYY-fs

AI + internet = AI + stolen

AI + Internet = Internet of Things, right?

@Ian, Yes but I think what 8Pla-net was referring to was that if one publishes anything creative or possibly useful on the Internet, consider it gone, taken. Stolen might be a harsh word actually because the moment one posts Anything on the Internet, not only is it There for all to see, it is There forever!

So, posted things, details, photos, information, theories, code samples, ideas, formulas, algorithms, anything and everything, is now available to everyone. What one might think is secure is only a challenge to others on the net.

A.I. + Internet = security = 4: privacy =0 (on a scale of 1 lowest, 10 highest)

Just my take

Don Patrick - Dec 29, 2016:

... Most of what I do in terms of virtual assisting is an exercise in futility, just to see if there’s any practical use for the A.I. . A minor benefit would be that I could include dynamic conditions like “Wake me 10 minutes before the Muppet Show starts”, a feature Google Now seems to have removed. But that too is nothing a dedicated scheduling app couldn’t do.

My current Cable provider allows me to use my remote to say, “SyFy Channel” or “Young and the Restless” (if my spouse), and the TV will go to that selection.

Otherwise if I say, “Wisdom of the Crowd”, and it is not time for it, the TV will alert me to that fact and allow me to set a time for a notification from the TV that my selection will start in 15 minutes (or other possible times I could have set).

These devices are getting smarter all the time.

I think it’s interesting to note that providers like Netflix will sent a message to the user / consumer that they have just added a movie they think I might like to their lineup of movies.

Obviously they are going off of some of my former preferences and making an “intelligent” decision based on those likes of mine. Still I sure some A.I. selective process is at hand somewhere.

Interesting, I didn’t know that functionality was already available. I suppose some things are easily improved with the digitization of television.

Don asked the summation, “AI + internet = ?”

So, let’s use sigma…

?
∑ AI.
AI=Internet

Art answered, “gone”

gone
∑ AI.
AI=Internet

I think “gone” is the best!