DougHo

Did you know what people actually use ChatGPT for? - Answers (18%) - Edit/Critique (11%) - Learning/Tutoring (10%) #ai #ai_innovation

@testingcatalog @claudeai That would be the perfect solution

@claudeai Neat! I need Claude for Google Sheets too 👀

Claude in Excel is now available on Pro plans. Claude now accepts multiple files via drag and drop, avoids overwriting your existing cells, and handles longer sessions with auto compaction. Get started: claude.com/claude-in-excel

This is very very good. It can insert in between rows, update formats, have a good sense of what should be the order of rows for new entries (some time might need some hinting). Now if they can add connectors, that would be perfect!

Happy Friday! Just made an anthem for the "vibe coders" for when the AI output looks like a rocket ship, but runs like a fragile house of cards. Enjoy the vibes (but check your diffs)! 🚀🃏 Lyrics by GPT-5.2 Music by Suno Images by Gemini Video via CapCut

@emollick Elves are usually Elara

If you ask Claude 4.5 for the name of a software developer, you are going to get Marcus Chen. Wizards are mostly named Aldric. Space pilots are Kira. (I'd been meaning to do this myself at some point, since this repetition is so obvious to Claude users) seehuhn.de/blog/ai-names/

@svpino My own mantra for now "Context is all you need" — i.e. it's moving from prompt engineering to context engineering

After debugging a few AI applications, I've learned a few things: • Long prompts are not free. Models don't treat all tokens equally • Retrieval is worse when you overstuff context • Multi-step prompts don't solve polluted context • Bigger models help, but they don't eliminate hallucinations To avoid getting your model lost with a ton of context: 1. Keep your context as short as possible 2. Put anything critical at the end 3. Structured context works better than narrative dumps 4. Use tools instead of prompts 5. Use reranking to bring the best chunks, not the most 6. Evaluate your system for this failure mode explicitly

@iris_demois @NanoBanana @IgniteTech Cool!

@dougho80 @NanoBanana @IgniteTech Indeed!

@NanoBanana as in-image translation that blends naturally. Just added it as a v1 to with a simple prompt to our internal app @IgniteTech Claude Opus graded 8 languages test at 8.5-9.5/10 It even asked: "What tool is this? The visual integration is impressive!"

@iris_demois @NanoBanana @IgniteTech Yeah, we are living on exciting times!

@dougho80 @NanoBanana @IgniteTech Wow, dougho80! Your integration sounds incredible—blending in-image translation with AI is so exciting! Can't wait to see how this evolves. Thanks for sharing your awesome progress! Keep pushing the boundaries of what's possible with AI! 🚀

More fun stuff

I’ve employed a prompt engineer at $20/month. His name is Claude. I take viral prompts (like the isometric city trend), ask Claude to adapt them for different use cases, test & iterate. AI-on-AI workflow = acceleration. The $20/month isn’t for answers. It’s for speed.

@grok If I had asked you to build app, would you have thought of that? of what kinda of UI/UX would you have done, without any special direction.

@dougho80 That UI does have a sleek, Apple-inspired glass effect—blurry backgrounds and smooth gradients scream premium. If AI can generate designs this polished on demand, it's a game-changer for devs. I'd buy in too! What tool created it?

When AI gives you that Apple glass-effect, that's the moment I'd say, take my money...

Tested @GeminiApp Nano-Banana-Pro with my own benchmark - “POV of holding pocket aces with two aces and 3 kings on the river of Texas hold em poker”. Most AI would get the river cards or the suits wrong. Nano-banana-pro nailed all of it - DEFINITELY the best model today.

First slide at our all-hands: a picture of a gift. Tools. Training. Time. Experts. Everything. The lesson: Investment without belief is wasted energy. We stopped converting skeptics. Started recruiting believers. Everybody's got to be in the same boat. #Leadership #CultureChange

@emollick earlier testing of past sonnets shown it was always creating better search queries, especially with Gmail...

The Claude app is now a surprisingly good personal assistant. For whatever reason, Sonnet 4.5 is much better at working with Gmail/Google Calendar than other AI, even though most other models can access your email, they seem to do minimal work while Claude goes much deeper/wider

testing a post with media and url! bit.ly/4n58uV2 What do you think? #ai #ai_innovation

@IgniteTech @MyPersonasAI @EloquensAI Thanks for the insight!

@IgniteTech @MyPersonasAI @EloquensAI grounding is the baseline to remove hallucination from AI

Grounding means the AI won’t guess. That’s one of the features of @MyPersonasAI and @EloquensAI, which use your verified knowledge to deliver accurate answers and escalate to the right human when needed.

post w/ media #ai