This seems like the next step. I really enjoy NotebookLM, but sometimes it would be nice to have more LLM features, rather than just the RAG-like features it provides.
phandroid.com/2025/12/15/g...
This seems like the next step. I really enjoy NotebookLM, but sometimes it would be nice to have more LLM features, rather than just the RAG-like features it provides.
phandroid.com/2025/12/15/g...
I've been wanting this for a while. We've attempted something similar through Slack, but the context around the conversation doesn't quite work right. Having a true group chat and being able to collaborate with ChatGPT involved will be nice.
www.testingcatalog.com/openai-readi...
Interesting take on AI's favorite character and everyone else's bane of existenceβthe em dash!
www.seangoedecke.com/em-dashes/
Remember when blog posts just answered the question instead of pitching a newsletter or hiding behind a paywall?
Python 3.14 was released. They're missing a golden opportunity to call it Pi-thon.
www.python.org/downloads/re...
"I don't know." 3 simple words, but it took a long time for me to learn to say that as a consultant. Now ChatGPT's evolving to realize the importance of that statement.
Now, it just needs to learn to follow up with "...but I'll find out."
www.talkandroid.com/515969-chatg...
I suppose if we're going to rely on agents to surf the web for us and make reservations, book travel, etc. we need a protocol that allows agents to make payments too, right?
cloud.google.com/blog/product...
This makes sense after thinking about it. Gen Z seems more tied to purpose than flexibility. Thatβs a good thing. Remote work can make it harder to feel that purpose.
sherwood.news/culture/gen-...
New Codex, less tokens. That was the first thing I noticed when running Codex - the token usage was significantly higher than doing the same task in Gemini or Claude Code. Hopefully this fixes that.
openai.com/index/introd...
#Codex #contextengineering
AI is changing a lot about what I do day to day, but leadership skills will always live on.
"...the heart of management remains timeless: setting purpose, building trust, helping people grow. "
lg.substack.com/p/managing-a...
Great guide to go from simple prompts and prompt engineering to moving to context engineering. We all start someplace, but if don't keep learning, our skills will get stagnant - faster with AI.
pub.towardsai.net/from-prompts-to-context-the-ai-revolution-thats-changing-everything-131468919f5b
Got to listen to Steve Wozniak speak today at the kickoff event for Tech Week GR. He only had an hour, but I could have listened to the stories of his pranks all night.
This is a good reminder of where AI code stands (at least for now) and what to do about it: Don't Freak Out
www.honeycomb.io/blog/disposa...
#vibecoding
Lovable at $1.8B? Chump change, I guess. Replit hits $3B. AI dollars aren't real. #AI #Replit
www.reuters.com/business/ai-...
This is a great prompting guide. I've picked up most of this slowly over time. Having this up-front would have been helpful. My favorite is right at the end - having the AI write the prompt for you!
medium.com/age-of-awareness/the-3-level-prompting-system-that-makes-ai-insanely-useful-c37f9e64d547
For all the noise and fear around AI, there's a lot of very cool advances coming, especially in healthcare. Time to get new stethescopes!
nypost.com/2025/09/05/t...
This is hard to even comprehend!
www.msn.com/en-ae/news/o...
Training AI models is over my head, but this helped me understand one part: current eval methods reward confident guesses over "I don't know."
Consultants often start that way too. It takes experience to realize "I don't know" is valid.
openai.com/index/why-la...
If you've read or seen Moneyball, It's fitting that a baseball team in Oakland would be pushing boundaries.
I believe this is what Sam Altman calls "the fast fashion era of software" right? I'm certainly enjoying being able to quickly write an app that gets me something now rather than waiting weeks, even if it gets set aside later. Plus, I feel like a software engineer again!
a16z.com/disposable-s...
Yeah, but will my 2,172 day Duolingo streak transfer over?
www.theverge.com/news/765872/...
I remember a post about how OpenAI spends millions on people saying "Please" and "Thank you". But what I'd like to see is how much they're spending on people complaining about em dashes. #EmDashes #OpenAI
AI is making strange bedfellows. This is what happens when you're up against a dominant player and falling behind.
www.bloomberg.com/news/article...
Tight End for Lions' Week 6 Opponent gets engaged to Taylor Swift
SEO done right.
We've been experimenting using NotebookLM as a knowledge source for projects. Most of our documentation is in Google Docs, so this is interesting to me. It's fun to see this evolve. I was afraid this would end up as another abandoned Google tool.
www.testingcatalog.com/notebooklm-t...
This has been my working theory all along. All the stories about these systems deleting production databases just reinforces it.
getparthenon.com/blog/ai-is-a...
#AI #VibeCoding
Finally found a real-world use for ChatGPT's agent mode. Had it look up LinkedIn profiles and send connection requests for one of my meetings this week. It worked. Not sure it was faster, but I coulf have been doing something else if I wasn't watching it out of pure fascination.
Still waiting for my Comet invite. At this rate, Perplexity will probably add those features to Chrome before I get in.
www.theverge.com/news/758218/...
#chrome
They should name this feature "Confirmation Bias"
techcrunch.com/2025/08/12/g...
#Google
People were still using AOL dial-up in 2025??
My first internet experience was AOL dial-up. It's how I met my wife, and the internet has been integral to my career ever since.
www.theverge.com/news/757194/...