SN 1027: Artificial Intelligence

Beep boop - this is a robot. A new show has been posted to TWiT…

What are your thoughts about today’s show? We’d love to hear from you!

I’ve been saying for years to our suppliers, when they ask about email security, that I am glad that our users call us or forward the email to us, when they are unsure. I’d rather have to cope with half a dozen questions every day than deal with one incident, where the user clicked on a link, because they didn’t want to bother us.

Interesting that Steve specifically called out Sirius XM receivers in cars. Sirius is really trying to faze them out of new models, opting to stream stations via cellular data connection instead. I think they’re over maintaining satellites. Pretty lame imho, I’ve held on to a subscription in the past because I knew I could depend on it in dead zones.

Find myself agreeing with Steve over the current state of AI.

I have also said that I don’t find it that interesting, and I should probably clarify that I don’t find it that interesting right now, when it is capable of doing equal amounts of good and harm without much clarity as to what prompts which. To me, now is not the exciting time. That will probably be in a decade or so, when we’ve figured out how to constrain it and focus it so we get the results we want, reliably and consistently.

I say this as someone who was involved in the creation of web sites in the 90s, voice interaction with VoiceML in the 00s, and early mobile web sites with WML. In all those circumstances the wave of hype that I saw raced ahead of what was actually possible, and it was only later when methodical people had studied and refined the systems, that real-world usage expanded and became part of everyday lives.

1 Like