20B-parameter Alexa model sets new marks in few-shot learning

Re:20B-parameter Alexa model sets new marks in few-shot lear... by omega3 on Tuesday August 02, 2022 @06:50PM (#32325282)

> This is much more complex, because you are asking the application to setup some sort of workflow - after it understands what you want it to do, it then has to work out how to execute that, which will be utilising the device API's / services.

I don't see conceptually the difference between the first and the second example. You're still executing two distinct actions, first being the waiting for x amount of time?

[ Parent | on HN ]

Re:20B-parameter Alexa model sets new marks in few-shot lear... by dietr1ch on Tuesday August 02, 2022 @07:47PM (#32325821)

The deferred execution makes this more complex as now the yet to be executed part needs to be queued somewhere

It could be issuing a command in the same way you do the first, but probably most lights just support switching them on/off right now, and won't take a delay argument.

So, if you can't just do the same you were doing, how do you do it? If there's some kind of local support, you can issue a delay command to some device in the network that will switch the lights on, but you still need to be able to control that action as the user might want to cancel or adjust the delay too.

If the execution need to happen remotely because there's no local support, then you don't even know that you'll be able to reach the device in 5 minutes (Internet down, router not letting random connections in). And keeping this request queued needs some infra in the server side too together with the necessary APIs to allow adjusting and cancelling requests.

[ Parent | on HN ]

Re:20B-parameter Alexa model sets new marks in few-shot lear... by Shorel on Wednesday August 03, 2022 @05:45AM (#32329604)

Just put everything in queues, even immediate things.

The AI feeds the action queues, and another isolated component performs the actions, including delays and follow ups.

[ Parent | on HN ]

Re:20B-parameter Alexa model sets new marks in few-shot lear... by shepherdjerred on Wednesday August 03, 2022 @01:01AM (#32327832)

> If there's some kind of local support, you can issue a delay command to some device in the network that will switch the lights on, but you still need to be able to control that action as the user might want to cancel or adjust the delay too.

I'm pretty sure there is some form of local support, at least for Google. Whne I last used it you had to choose a speaker that automations executed from

[ Parent | on HN ]

Re:20B-parameter Alexa model sets new marks in few-shot lear... by hmryehbut on Tuesday August 02, 2022 @08:06PM (#32325978)

So it requires things software engineers do all the time.

Where is the hard part?

[ Parent | on HN ]

Re:20B-parameter Alexa model sets new marks in few-shot lear... by NLPlatypus on Tuesday August 02, 2022 @08:37PM (#32326233)

But you can apparently tell it “remember I put my keys upstairs”, and when you ask where your keys are, it tells you upstairs.

There is storage going on already

[ Parent | on HN ]

Re:20B-parameter Alexa model sets new marks in few-shot lear... by eru on Tuesday August 02, 2022 @10:52PM (#32327093)

> You're still executing two distinct actions, first being the waiting for x amount of time?

It's not a fixed amount of time. What if you rewind in the song a bit to catch something you missed in the lyrics?

[ Parent | on HN ]

Re:20B-parameter Alexa model sets new marks in few-shot lear... by ajnin on Wednesday August 03, 2022 @05:32AM (#32329509)

It's not distinct, the second action depends on the first one being finished.

I think this highlights that AI assistants are not really "I", they are good at doing fuzzy, hard to specify things like understanding speech, but there must still be an engine behind that interprets the text and transforms it into execution steps, and that must still be done by developers so it will be limited.

[ Parent | on HN ]