MAUREEN

@FellMentKE Feels like we're missing a memory layer for AI development.

AI coding has a weird habit. It keeps solving the same problem twice. Not because the model is bad. Because every new session acts like yesterday never happened. 👇

@FellMentKE The jump in quality from older AI 3D tools is becoming obvious.

What if your 3D Model generator actually thought before building like an LLM for geometry? Meet Rodin Gen-2.5, Its the world’s first 3D AI that delivers 10M+ polygon models with skin-level microstructures… You can generate 1M polygons in just 4 seconds. Check The Thread👇👇

@Ambani_Wessley Adding Micro shows how rapidly AI assistants are advancing.

Me adding Micro into my list of AI Assistants and it overpowers all of them including Claude, Codex & OpenClaw!

@kate_osita_ Being able to direct the creative process matters a lot.

Your wildest story, a cinematic reality? TapNow Agent, with pro toolkits & Seedance 2.0, makes it happen. Transform your inner world into an episodic IP. Enter the 10,000 Parallel Universes contest & direct your destiny today! app.tapnow.ai/tapflow/view/f…

@dee_naliaks Himalayas is underrated and deserves more attention from job seekers.

Some remote jobs only operate between 1:30 AM and 6:30 AM, which is why hardly anyone applies. I’ve identified overnight remote jobs paying up to $200 per hour in USD. Here’s each job and why they pay so much👇

@FellMentKE Remote work changed my life so I always save threads like this.

STOP SCROLLING PAST REMOTE JOB POSTS. SOME OF THE BEST OPPORTUNITIES NEVER GO VIRAL. HERE ARE 10 REMOTE JOBS YOU SHOULD CHECK OUT TODAY 👇

You're ONLY BROKE & it's TEMPORARY, you're NOT POOR, don't be depressed or pressured.

Being a successful adult is not about knowing everything. It is about knowing how to find answers and not being afraid to try and fail.

ADULTS ONLY OPEN THE THREAD !!!!!....

POV : Avoid old women The women :

@dee_naliaks The current forget and repeat workflow is frustrating.

AI coding agents have one expensive habit: They forget everything. Every time a bug comes back, the agent starts another investigation like it's seeing the codebase for the first time. I tested a different workflow using EvoMap and the results were surprisingly practical 👇

@FellMentKE This is the kind of AI story that flies under the radar at first.

🚨 While everyone is focused on OpenAI, Google, and Meta... Russia might be preparing an AI device that nobody saw coming. 👀 Reports from the St. Petersburg International Economic Forum suggest that Sber is developing a new AI-powered hardware product. 🧵

@FellMentKE Definitely grabbing the credits and testing this myself.

hot take: every voice AI product you've used has been built on a transcript. not the audio. a transcript of it. @modulate_ai just shipped a model trained on 550M hours of raw audio that skips that step entirely. 150+ behaviors detected. real-time. out of the box. it's called Velma and it's the primitive voice products have been missing. essentially, the model powering the voice pipelines of COD and GTA onine is now available as an API. they're giving the first 1000 credits out for free.

@n__deborah @CallofDuty @GTAOnline @modulate_ai Most people underestimate how hard voice intelligence actually is.

the voice intelligence running inside @CallofDuty and @GTAOnline just became available to every developer via API. same model. same production scale. @modulate_ai didn't build a demo and work backwards. its called velma built by @modulate_ai What got my attention here is that Velma wasn’t built as a demo-first API. Velma has been running inside some of the world's largest voice environments before they opened it up. the difference from every other voice API: it doesn't transcribe your audio and hand it to an LLM. it listens. tone, hesitation, intent, emotion - all of it, in real time. this is the one worth paying attention to. The world FIRST audio-native voice AI 1,000 free credits to try it: modulate.ai/api/velma?utm_…

@dee_naliaks @ElsieEssie_ If the numbers are accurate, that's a pretty bold statement.

🚨 this startup just posted the most uncomfortable benchmark chart in voice AI. velma - their model sits top-left on the conversation understanding benchmark. beats GPT-5. beats Gemini 3 Pro. beats Grok 4. 🤯 at a lower cost than any of them. the company is called Modulate. barely anyone's talking about them yet. they just released their API. the one powering COD and GTA online. 1,000 free credits for developers: modulate.ai/api/velma?utm_…

@kate_osita_ Most meaning is hidden in tone, timing, and delivery.

speech-to-text → classify the transcript. that's been "voice AI" for a decade. @modulate_ai looked at that stack and built something completely different. they just opened up the world's FIRST and ONLY audio-native voice intelligence model to the world!!! Velma doesn't read your audio. it understands it. hesitation, sarcasm, emotion, vocal stress - 150+ behaviors, straight from the signal. no transcript in the middle. this is the more native way to build voice products. took someone long enough. their 1,000 API credits are free: modulate.ai/api/velma?utm_…