I just tried xAI's new Grok Voice Agent Builder, and my first impression is that it's taking a very different approach from other voice agent platforms. Compared to platforms like Vapi, Retell, and Bland AI: • Built-in real-time web search – No need to integrate external search APIs for up-to-date information. • End-to-end stack – Telephony, voice model, reasoning, and tools are all integrated into one platform. • Sub-second responses – Conversations feel fast and natural. • Simple setup – You can build and deploy a voice agent in minutes. • Competitive pricing – Starting at $0.05/minute. Platforms like Vapi and Retell still have advantages in terms of ecosystem maturity, integrations, and production deployments. But Grok's biggest strength is its all-in-one experience and native access to real-time information. The AI voice agent space is moving fast, and Grok just became a serious competitor. What matters most to you in a voice agent platform - speed, integrations, pricing, or built-in search?

Seedance 2.0 dropped and it's not just another video model. Paired with Claude for scripting, teams are generating full "creator-style" product ads — hook, demo, CTA — without booking a single human creator. Here's what changed: → Native audio-video sync — no separate voiceover pass, dialogue and sound generate together → Multi-shot consistency — same "person" holds the same face, product, lighting across every cut → Speed — a batch of hook variants in the time it used to take to brief one creator I've watched founders in our builds go from "we need 10 UGC creators for testing" to "we need one good product photo and a prompt." But here's the part most people skip: this doesn't kill UGC. It kills slow UGC. Real customer trust, real testimonials, real founder-led content — that's not going anywhere. What's disappearing is the 2-week turnaround for a simple hook test. The founders who win the next 12 months won't be the ones who ban AI creative or the ones who go all-in blind. They'll be the ones who know exactly which ads need a real human and which ones just need to ship fast and get tested. Where do you sit on this? Are you testing AI UGC yet, or still fully on real creators? Drop a comment — curious how many of you have actually run a Seedance-generated ad live.

Google just released Nano Banana 2 Lite, a faster, cheaper image generation model. Creates images in ~4 seconds at $0.034 per image. Rolling out across Search, Gemini app, Photos, NotebookLM, and more. Also expanding access to Gemini Omni Flash, their multimodal video editing model, turn text/images/video into short video clips (up to 10 sec) using natural language prompts. $0.10 per second of video. 🔗 Bonus: you can chain them, generate an image with Nano Banana, then animate it with Omni Flash. Both come with SynthID watermarking for transparency on AI-generated content. Big week for AI updates 👀

Hi all. I've just spent another 24 hours going around in circles trying to solve a routing/mapping problem. I partly 'fixed' my problem but not the bit that really matters! I've run repeated tests but now I'm just confused and consuming credits!!! Can you help? The screenshots show what the output looks like when more than one URL is found, and what happens when one URL is missing. So the challenge is how to use the upstream information to 'pretend' the missing URL exists. This is the test scenario that can be used for re-runs without affecting live data: - 🧪 [TESTING] GSC-Existing Content Scan (v2.1) - https://us2.make.com/93886/scenarios/5554137/edit

I've been exploring different workflows lately, and it's amazing how the right combination of tools can completely change a process and save hours of manual work. Curious to know which app integration has had the biggest impact on your workflows and what made it so useful.