#Veo3 is strong. Nice job @GeminiApp @GoogleAI

Pretty good!. #NewMusicFriday #newmusic #CountryMusic #folk.

#SoftwareDevelopment talent is flipping. Instead of a resume and certs, candidates will bring orchestrated agent packs and eval scores. BYS (“bring-your-swarm”) becomes the hiring advantage. Think I’m wrong? tell me why. #ai

Just boarded the brand new, beautiful-and huge-#staroftheseas for the naming ceremony sailing with @RoyalCaribbean!

Horse racing #LLMs to get aligned output is useful, but both @claudeai #Opus and #GPT5’s conclusions at the end of my last experiment made me laugh when I asked who should lead the scoping work. Neither one wanted to do it. Can you imagine if this were real people talking

What #ai models need is a micro-nudge capability when in thinking mode, or even mid-output. Thumbs up / thumbs down buttons to signal the model is on the right track. Thumbs down and it pauses and prompts for a clarification. Heck, even a pause button would suffice. Yes, I know.

I think this captures most people's experience with #vibecoding.

What a diff a week makes. Been using #GPT5 (high fast max) for the last few days in @cursor_ai. It’s been crushing it. If you ditched gpt5 after a couple of days like I did initially, give it another shot.

I try to keep it positive, but this a bit of an odd flex.

Watching this investment manager talk about @Apple on @CNBC. General thesis I can buy, but I’m growing tired of the “apple isn’t always first, but they do it better” argument. I’m also not buying the argument that “the cost of intelligence is going to zero…” because of vibe.

I tried. Really. A weekend and today with @cursor_ai CLI and #GPT5. Back to #Sonnet tomorrow, but I may take a detour and finally give #Grok a run. GPT5 wasn't terrible, but it just couldn't ever seem to get where I wanted to go despite solid specs+prompts.

With all of the amazing #ai advancements, is anyone else waiting for the ai cage match video battle between @sama, @elonmusk and @sundarpichai? @elonmusk - I bet Grok could do it.

AI Model Rap Battle Round 3! #GPT5 vs #Gemini 2.5 Pro vs #claude Opus 4.1. I don't know if Claude Opus 4.1 or Gemini 2.5 Pro is the winner, but I think it's fair to say GPT-5 is not tuned for rap lyrics. Not sure it understood the assignment.

AI Model Rap Battle Round 2! #GPT5 vs #Gemini 2.5 Pro vs #claude Opus 4.1.

Who was the Round 1 #GPT5 vs #Gemini vs #Claude Rap Battle Winner? Listen here: #ArtificialIntelligence.

The ultimate benchmark test. #GPT5 vs #Claude vs #Gemini. @thisdayinai inspired me to produce a true 3-round rap battle where each model responds to the next. Who won Round 1? Rounds 2 and 3 are hilarious. Stay tuned.

Not loving #gpt5, at least not in @cursor_ai. It's hard to put my finger on it. It isn't bad, it's just not amazing. Over-hyped expectations? Maybe.

The @openai gaggle of nerds is having too much fun with this #GPT5 demo. I just hope we get some use cases that are novel. Credit to the @thisdayinai for the term.

This intro is an ai vid, pretty sure.

2 mins til GPT-5….