Attending ‘South-by’ with Otter.ai
Hi all. This week we’ll cover AI-assisted search, bias in AI, AI in news, prompt engineering, and finally a guy who wore an AVP at his wedding. But first…
Greetings from Austin!
The South by Southwest (SXSW) Interactive conference is wrapping up as I type this. A group of folks from the world of public media are here soaking up the tech trends and thinking about how it impacts our industry. The Public Media Innovators’ 3rd Thursday webinar, Innovation at SXSW: What Public Media Should Know, happens next Thursday, 3/21 at 1pET. Hosted by David Huppert from PBS North Carolina , the panel will feature, Melissa Thibault, Ph.D. from PBS NC, Mikey Centrella from PBS, and Katie Garrett from the Utah Education Network (and possibly others).
Attending ‘South-by’ with Otter.ai
One tool that I’ve been finding helpful here at South-by is Otter.ai. We first heard about Otter last summer in the context of being a meeting surrogate. Now, more than six months on, they are striving to be “Meeting GenAI”.
Ostensibly a transcription app, Otter layered an AI interface on top that saves you the need to take detailed notes of conference sessions. It records and transcribes the audio using your phone or laptop mic, and then stores that in the cloud.
Otter’s value-add here is the elegance of the interface, married to the usefulness of the AI tools. It isn’t a effortless solution, but for me, who sits in a lot of sessions here at SXSW, I found that instead of trying to capture every little detail, I could instead make fast notes about things I wanted to ‘ask the transcript later’. This let me focus on having ideas, or creating action items.
Two notes of caution. Otter does seems to require an internet connection, and while it doesn’t have to be robust, if you lose that connection it’ll end your recording (which carries a high P.I.T.A. factor with it). Additionally, the mic can pick up chatter from you or people sitting around you so - pro user tip - I found sitting near the PA system helps.
Outside of conferences, Zoom is offering this type of tool now for calls (where Otter works as well), as is Microsoft Teams. So this isn’t new tech. But it’s gaining popularity. In fact, Katie Garrett, from UEN, tells me that many students in Utah are using Otter for disability accommodations when they need assistance taking notes.
A Pro account is going to run $17/month or $124/year, but you can get a free 7-day trial if you register through the website. It might be worth trying the free version for a week if you are going to Public Media Venture Group 's TechConnect ‘24 or the 2024 PBS Annual Meeting.
Because I’m in Austin this week, the newsletter will be a little shorter. And though you’ll find references to South-by sprinkled throughout this week’s exploration, I’ll have full SXSW coverage next week. Okay, on to the links…
If You Click Only One…
AI-enhanced search tools are here — and public media needs to pay attention ( Jason Katterhenry - Current: News for Public Media ) - I was really excited to see this hit Current last week. Tools like Arc (covered by Jason, above) and Perplexity.ai (more on that below) have had open tabs in my browser for a few weeks. Naturally, they started popping up as topics of conversation around the conference here. I’ve only started to dabble with them a bit for vacation planning (always my test case), but if you are someone concerned about SEO at your organization this is an area of AI development that you need to watch carefully.
Things to Think About…
Gender Bias in AI (International Women's Day edition) (Yennie Jun - Art Fish Intelligence) - This line got me thinking: "Should language models reflect the biases that already exist in the real world to better represent reality?...Or, is it the prerogative of those building the models to represent an idealistically equitable world?" The question of "whose ideal" seems to be one plaguing many aspects of society these days. Regardless, this piece does offer a good literature review and discussion (skip to "Discussion" in her piece if you want to read the good stuff first), and provides a good reminder of how far we still have to go with bias in AI.
The big design freak-out: A generation of design leaders grapple with their future (Robert Fabricant - FastCompany) - South by Southwest always has a fairly robust design track for its sessions. But this year something felt different. A colleague and I went to an Ideo session because you used to be able to count on Ideo staff to give you some good big brain thoughts. Before going, another colleague said to us before the session, "Is Ideo even relevant any more?" And she was right. If you are in design, this may not be news to you, but for the rest of us it may put a finger on something you've sensed but can't name.
No one really knows how the next ten years are going to play out (Tony Vitillo - The Ghost Howls) - One of my early impressions of the SXSW conference this year is that people are really scrambling to make sense of what is happening across several aspects of emerging tech. And that's especially true with xR. So Vitillo's post, which was clearly written before South-by, resonated with me.
Recommended by LinkedIn
Inside the Music Industry’s High-Stakes A.I. Experiments (John Seabrook - The New Yorker) - David LeRoy sent this one to me earlier this year, and it took me the plane ride to Austin to finally have the time to read it. I love the extensive references to punk and post-punk, but ultimately this is a story about someone who has always been on the progressive edge of the music business, and through wisdom and experience is now trying to lead the way in an industry that - more than any other - has always been served by technology in the digital age.
Things to Know About…
AI in newsrooms: revolution or retooling? (Sarah Grevy Gotfredsen - Current) - About a month back I linked to a report in CJR about the use of AI in newsrooms. This interview with the author, Felix Simon, was a value add so I thought I'd revisit it here.
Teachers Are Using AI to Create New Worlds, Help Students with Homework, and Teach English (Ross Teixeira - The Markup) - You asked for more education coverage, and I do take requests. Here are seven examples from around the globe that will hopefully give you some insight for how educators are using generative AI.
Prompt library (Anthropic) - As I mentioned last week, Claude is back in play now that version 3 has been released. This past week, they released a prompt library that could be of interest to Claude power users.
Chain-of-Thought Prompting: It’s off the Chain (Daniel Nest - Why Try A.I.) - For those still interested in prompting, Daniel Test has this useful tip.
Midjourney debuts feature for generating consistent characters across multiple gen AI images (Carl Franzen - VentureBeat) - This may seem unimportant on the surface, but thus far every time you created a new image (even with the same prompt you, at best, got a variation on a theme. I mentioned the idea last week of creating station avatars. This step forward brings that idea much closer to being a practical reality.
Who are OpenAI's new board members? (Jaspreet Singh and Arsheeya Bajwa - Reuters) - I made some noise last fall in this newsletter about the fact that the OpenAI board had, following their failed coup, was all male. Now that they've rectified that, it's only fair to mention it here.
The State Of Play In Spatial Computing/XR In 2024 (Rony Abovitz via Medium) - A good overview of the xR industry from the found of Magic Leap (and speaker at one of our proto-3rd Thursday webinars back in 2021).
And finally…
We Talked to the Guy Who Wore a Vision Pro VR Headset at His Wedding (Maggie Harrison Dupré - Futurism) - And finally, hopefully the wife knows there's a 14-day return policy on that husband.
Have a creative, productive week!
Chad, remembering our many convenings at SXSW. Glad you’re keeping the tradition alive.
Co-Founder of Altrosyn and DIrector at CDTECH | Inventor | Manufacturer
9moIt's fascinating to hear about your experience with Otter.ai at SXSW. You talked about AI-assisted search and bias in AI; how do you envision overcoming biases in AI-assisted search algorithms to ensure fair and equitable results, especially in the context of news dissemination and prompt engineering? If you imagine a scenario where a news organization aims to eliminate biases in its AI-driven search engine, how would you technically approach this challenge for maximum effectiveness?