Hey folks, it’s JENNI with the AI.
It's time for a little introspection, a dash of retrospection, and a whole lot of YOU-direction. It’s JEN_AI and we’re looking back so we can spring forward.
This is JEN_AI Challenge 09: FEED_BACK
Get in the loop. Give FEEDBACK on JEN_AI
I got some helpful listener feedback that it was a bit hard to keep up with the JEN_AI outpourings - both in terms of understanding the mad ramblings and getting through the backlog of weekly episodes.
So this week- we’re hitting the pause button to reflect on the JEN_AI journey and dig out a few gems. I’m asking for reader/listener feedback and suggestions, and reflecting on when to take vs. ignore advice.
// Part 1: The JEN_AI Feed, Going Back
(a Reverse-Chronological Roundup of what we’ve learned so far)
In case you missed it - here’s a whirlwind rundown of JEN_AI’s experimental outputs and the key takeaways so far, flipped and summarized for your byte-size consumption:
#8 Build Something Good - we got AI to code a simple web app, tried to make a Twitter/X bot which got banned, digested government aid cuts and aimed for world betterment with special guests. We learned: Even AI needs a moral compass and maybe a lawyer. Try out: Replit to code websites and apps for you.
#7 Social Mediums - JEN_AI channeled social media, confronted harsh realities, gave some reader shout-outs and tried to read you. We learned: Keep it real- the true value is connecting with real living humans, duh. Use the tech, don’t let it use you. Don’t try: to get on all 20+ social media platforms.
#6 The Ultimate AI Challenge - We brainstormed useful uses of AI, including: brainstorming. We learned: Sometimes, more powerful than to speak is to listen. The process of refining your ideas back and forth with a chatbot can be more valuable than the generated outputs alone. Try it out with your fave chatbot e.g. ChatGPT, Claude or Google Gemini (links)
#5 Let’s Get Social - JEN_AI sent some Valentine’s Day love to listeners and invited them to get on the show. We learned: you can’t rush engagement, and wordplay is not overrated.
#4 AI Emergency! Deeply Seeking Shortcuts - The U.S. stock market and a stretched parent eyed-up DeepSeek - a "thrifty" language model - via an obscure 80s Madonna movie. We learned: Everyone's looking for a shortcut, especially when time, money and power is involved. Using a large-langugage model chatbot can help get that copy right.
#3 Future Podcast - One AI podcast guest predicted the future of tech while two others roasted us alive. We learned: AI-driven audio-quality is brilliant or brutal, with no in-between. Try out: Google’s NotebookLM to generate easy-to-digest podcasts about any source material and Descript to edit ‘real’ podcasts.
#2 Let There Be Speech - A free speech… about free speech… generated by not-so-free AI speech. A voice-cloned JEN_AI tried audio. We learned: Free speech can come at a cost (especially if you’re Trump’s speechwriter) and voice cloning is as creepy as it sounds. Try it? Or not.
#1 A NET_NEW Newsletter - We made an AI/human newsletter about writing an AI/human newsletter about the AI/human news. We learned: Chatbots can elevate your writing skills to the internet average, but that doesn’t make you a media pro.
#0 Intro to JEN_AI’s AI/Human Blog. AI/human writing about AI/human writing about AI. Try: use Substack as a media platform to publish blogs, newsletters, quick pics, podcasts and more. But maybe turn off its attempts to beg cash from your friends.
// Part 2: Feed the Machine
Time for some fine-tuning.
YOUR constructive complaints and suggestions can help make JEN_AI more useful, more entertaining, less irritating.
Please let me know (via this poll or any legal means) what you’d like to see more or less of, or any other feedback.
I’m also considering switching to sending the blogcast every 2 weeks instead of weekly - to be less spammy, to spend more time on things like video, the challenges themselves, & distilling the good bits for easy multi-channel consumption, and to give anyone interested a chance to keep up. Def let me know your suggestions!
Thanks for taking the time to make JEN_AI better!
// Part 3: Giving feedback on feedback
it wouldn’t be JEN_AI without unnecessary meta-content (or would it? Let me know). So this week I’m sharing a few notes on feedback.
I can distil some advice I got on giving tough personal feedback at work to this: be specific about something someone did and how that made YOU feel. Be respectful, don’t generalize, and reflect on how their actions impacted you personally. Behaviour can change, but not if you’ve dismissed their entire personality.
For some people, getting feedback can cause emotional distress so severe it feels like physical pain. This is sometimes referred to as Rejection Sensitive Dysphoria (RSD). Coaching or therapy can help with this, a lot, and is also something anyone can benefit from more generally.
The thing is- if we can’t take feedback, it’s hard to make progress or grow. And that’s whether you’re starting a startup or just trying to have a relationship.
Feedback works better in an environment of psychological safety and trust. (Not internet trolling.)
Feedback can be imperfectly expressed and might not even offer the best solution, but there’s a signal in there, a real human feeling, and someone who cares enough to want to help us learn.
So can JEN_AI get better with feedback? Alex Ramirez, a listener who is building a biotech startup called Antibody Design, kindly sent me some thoughts on the DeepSeek episode:
“The beginning was hard to focus on and I totally didn’t get the obscure movie parallel. Sounded like a mix of low energy scripted bits, mixed with high energy ramblings and chuckles. The bit with AI agents chatting [the Google NotebookLM generated podcast] was informative and easy listening. Maybe do a shorter intro and then the chatbots.
Feel free to ignore everything though! 😅 I get people ALL the time saying “maybe you should do this / change the product/ market/ strategy, etc” , and I ignore 90%. Just follow your gut and create something you like / would enjoy listening to.”
Alex, thanks for the excellent advice that I have failed to follow. To honour it a little bit I’ve attempted to reduce and clarify the rambling and references and I’ve added a NotebookLM generated chatbot podcast at the end explaining how AI neural networks use feedback to learn. More informative, easy listening. But I’m afraid that- for at least today- the irrelevant movie references and lame-pun laughter stays.
There’s a balance to be found when creating something you haven’t really seen before, of course there is. There’s a middle ground. But that’s the question - how far to go? And how to avoid making something old-hat or average, when what you’re really looking for is a new medium.
The answer? How to create psychological safety and improve? I say - experiment. Iterate. Learn. Learn how to listen to the opinions of experts, consumers, users, supporters and those who care, while still being guided by your own Inner Light.
// JEN_AI Playlist: This Week’s Theme Song(s)
Inner Light by Elderbrook + Bob Moses. And as a lapsed circus aerialist, I love the music video’s dance performance on a rotating platform.
With the AI (I mean, the IE) by the epic JENNIE Kim from BLACKPINK. Bonus points for sampling J-Lo’s Jenny from the Block. This lady knows how to spring forward while nodding back. We’re on the way up.
// References
Welcome to JEN_AI, where you should probably feel better about yourself if you DON’T get the references. But if you want to check the big ones, here goes:
FACE/OFF the movie: Face/Off is a brilliant, terrible sci-fi action film starring Nicolas Cage and John Travolta and directed by John Woo. One is an FBI agent, one is a terrorist, they swap faces, morals and identities. Cue action. Why even mention it when I’ve barely referenced it? Why the hell not.
“Jenni with the AI” riffs on “JENNIE with the IE”. See JENNIE’s music above.
There’s a lot I could attempt to say about how neural networks use feedback to learn, but I’ll leave that to the bots. I’ve generated another NotebookLM audio overview you can feel free to listen to at the end, and it contains all you need to bluff being an AI expert, or offers something to fall asleep to - you choose. Here it is:
An AI-generated AI-primer podcast made with Google Gemini and NotebookLM:
Thanks for listening! All views expressed are my own (or a chatbot’s, duh)
Share this post