meta_pixel
Tapesearch Logo
Log in
Let's Know Things

ChatGPT

Let's Know Things

Colin Wright

News Commentary, News

4.8593 Ratings

🗓️ 20 December 2022

⏱️ 31 minutes

🧾️ Download transcript

Summary

This week we talk about OpenAI, DALL-E 2, and voice assistants.

We also discuss snake oil, AI, and GPT-4.

Show notes / transcript: https://letsknowthings.com/episode343



This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit letsknowthings.substack.com/subscribe

Transcript

Click on a timestamp to play from that location

0:00.0

A few weeks ago, I did an episode on voice assistants, which are software offerings optimized to parse human language

0:23.3

as it's spoken, or in some cases as it's typed out.

0:26.9

And the primary innovation behind those applications is allowing them to be competent listeners,

0:33.0

and then reaching out into the ether, the internet, or sometimes offerings contained within other apps owned by the voice assistance owners,

0:41.3

and then summoning some relevant information or triggering relevant tasks when asked to do so by the user.

0:48.3

So if I'm using Amazon's Alexa voice assistant, I might speak within range of an echo device's microphones, a request

0:56.5

to set a timer for 20 minutes, and the voice assistant will then confirm that request, set a timer,

1:01.9

and in 20 minutes let me know that the timer has reached zero.

1:06.4

Similarly, if I'm using Apple's Siri voice assistant, I might speak a request within range of my iPhone's microphone

1:13.4

that I'd like to send a text to a friend, and I would speak that text, including each bit of

1:18.8

punctuation, and then say send, and that friend then receives a transcribed text version of the message

1:25.4

I spoke out loud. These are, according to the data we have on such things, fairly useful tools to have on hand.

1:34.3

Voice assistants have yet to become the ad placement paradise, or voice-activated buying stuff

1:40.3

stimulant that many of the companies behind them had hoped for. Few people are using them to buy things on Amazon,

1:48.0

and fewer willing to tolerate an abundance of ads in between starting timers or sending transcribed text messages.

1:55.0

But they are popular for those relatively simple, practical use cases, probably because these use cases allow us to utilize

2:03.1

existing technologies in new ways, which is similar to what smartphones did for existing technologies

2:10.2

that we already had on our laptops and desktop computers.

2:14.4

They're a repackaging of what was already there, which in turn makes existing tools

2:19.9

more useful or just differently useful because of the new context in which those tools can be

2:25.9

leveraged. And because of that new convenience and the intuitive access they enable, it also

2:31.5

makes these tools available to more people, which is also important.

...

Please login to see the full transcript.

Disclaimer: The podcast and artwork embedded on this page are from Colin Wright, and are the property of its owner and not affiliated with or endorsed by Tapesearch.

Generated transcripts are the property of Colin Wright and are distributed freely under the Fair Use doctrine. Transcripts generated by Tapesearch are not guaranteed to be accurate.

Copyright © Tapesearch 2026.