Hackaday Podcast 218: Open Source AI, The Rescue Of Salyut 7, The Homework Machine

This week, Editor-in-Chief Elliot Williams and Kristina Panos have much in the way of Hackaday news — the Op Amp Challenge is about halfway over, and there are roughly three weeks left in the Assistive Tech challenge of the 2023 Hackaday Prize. Show us what you’ve got on the analog front, and then see what you can do to help people with disabilities to live better lives!

Kristina is still striking out on What’s That Sound, which this week honestly sounded much more horrendous and mechanical than the thing it actually is. Then it’s on to the hacks, beginning with the we-told-you-so that even Google believes that open source AI will out-compete both Google’s own AI and the questionably-named OpenAI.

From there we take a look at a light-up breadboard, listen to some magnetite music, and look inside a pair of smart sunglasses. Finally, we talk cars, beginning with the bleeding edge of driver-less. Then we go back in time to discuss in-vehicle record players of the late 1950s.

Check out the links below if you want to follow along, and as always, tell us what you think about this episode in  the comments!

Download and savor at your leisure.

Episode 218 Show Notes:

News:

What’s that Sound?

Interesting Hacks of the Week:

Quick Hacks:

Can’t-Miss Articles:

7 thoughts on “Hackaday Podcast 218: Open Source AI, The Rescue Of Salyut 7, The Homework Machine

  1. What threw me off with the percolator sound is that it was too fast and consistent. The percolators I’ve used had a significant and random delay with the bubbling cycles, which is why I though it was something that was drawn back and forth by manual force like some rattling bow mechanism.

    Re. standing in front of a self-driving car so it would notice a person: that’s not guaranteed – with example cases out in the wild where the AI just thought “There appears to be a person in the way, but they’re not supposed to be there, or this is a false inference, so this is not a situation I need to react to. I can just drive on.”. You risk the AI getting confused by circumstances or just not reacting appropriately even when it does perceive a person and… running you over. Would you stand in front of the car to stop it? Do you trust the programming enough? I wouldn’t, for at least the next 20 years.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.