DeepSeek V3.1 Just Arrived

Welcome to another edition of AI PlanetX.

DeepSeek V3.1 redefines open-source AI; Pixel 10 showcases AI-first design; Meta forms Superintelligence Labs under Alexandr Wang.

• Hottest AI News

• Top AI & SaaS Tools

• AI Tutorial: Generate Video Subtitles

• Top AI & Tech News

• AI Art Spotlight

• Prompt of the Day: Humanize AI Writing

• AI Video Tutorial

• Course of the Day: LLM from Scratch – Code & Train Qwen 3

Hottest AI News

High-Flyer

DeepSeek V3.1 Just Dropped and Could Be the Most Powerful Open AI Yet

DeepSeek just dropped V3.1, a 685B parameter model that’s turning heads in the AI world. Quietly launched on Hugging Face, it’s already outperforming top proprietary systems — while staying fully open and available to all.

Details:

• DeepSeek V3.1 hit 71.6% on the Aider benchmark, matching Claude Opus while costing 68 times less—1.01 per task vs 70. It handles 128k tokens and unifies chat, reasoning, and coding in one system, fixing issues other multi-capability models face

• Released just weeks after GPT-5 and Claude 4, it contrasts US firms’ costly APIs by offering F-R-E-E download and modification. It supports multiple precision formats plus tokens for real-time web use and internal reasoning

• This underscores a divide: US firms guard AI as IP, while Chinese groups push it as a public good to speed innovation. DeepSeek’s approach reduces fragmentation, with savings worth millions for enterprises

DeepSeek V3.1 may mark the point where open-source systems catch up to proprietary ones. The key question is who will control the future development and distribution of the most advanced AI.

Create How-to Videos in Seconds with AI

Stop wasting time on repetitive explanations. Guidde’s AI creates stunning video guides in seconds—11x faster.

• Turn boring docs into visual masterpieces

• Save hours with AI-powered automation

• Share or embed your guide anywhere

How it works: Click capture on the browser extension, and Guidde auto-generates step-by-step video guides with visuals, voiceover, and a call to action.

Best part? It’s 100% free.

👉 Download the extension now

DeepMind

Google’s Pixel 10 Reshapes What a Smartphone Can Do With AI

Google just released the Pixel 10 series, and it's clear they're not playing around when it comes to AI integration. While Apple is still fine-tuning cameras and picking colors for the iPhone 17, Google is literally reimagining what a smartphone can do.

Details:

• Visual Overlays might be the coolest addition. Point your camera at a foreign street sign and get real-time guidance overlaid on your screen. A lifesaver for travelers lost in translation

• Magic Cue revives the proactive assistant vibes of Google, but with AI power. It suggests restaurants while you text, helps with reservations, and reminds you about that online return you keep forgetting

• Camera Coach could change photography for enthusiasts. The AI suggests better framing in real time, while "Auto Best Take" scans up to 150 shots to deliver the perfect group photo

The Real Game-Changer:

• Voice Translate for phone calls might be the most impressive feature yet. Real-time translation that maintains each speaker's voice across 11 languages? If it works as advertised, this could revolutionize international business and travel.

Google isn't just adding AI features – they're fundamentally changing how we interact with our phones. While we wait until 2026 for Apple's AI Siri, Google is already delivering a glimpse of the truly intelligent smartphone future.

Top AI & SaaS Tools

• Dialora (Life-time Deal): AI voice agents that answer calls, qualify leads, call back promptly, and book appointments using your business-trained, on-brand responses

• Firecrawl v2: Scrape websites 10× quicker (~1s), guide it using natural language (e.g., "fetch product pages"), and automatically extract news, images, and summaries [F-R-E-E]

• Google Whisk: Create and remix images by dragging and dropping photos as subject, scene, and style inputs, available in 77 countries [F-R-E-E]

• LIA-X: Portrait animator that copies facial motion from a video to a photo using interpretable motion codes for precise, controllable edits [F-R-E-E]

• AI Sheets: Create datasets in a spreadsheet — describe the data you need and let AI populate the entries with no coding required [F-R-E-E]

Master ChatGPT for Work Success

ChatGPT is revolutionizing how we work, but most people barely scratch the surface. Subscribe to Mindstream for free and unlock 5 essential resources including templates, workflows, and expert strategies for 2025. Whether you're writing emails, analyzing data, or streamlining tasks, this bundle shows you exactly how to save hours every week.

Subscribe for Your Free Bundle

AI Tutorial

How to Generate Timestamped Video Subtitles with Gemini 2.5 Pro

Creating subtitles in multiple languages can be time-consuming — but with Gemini, it becomes a much smoother process. This step-by-step guide will show you how to generate accurate, timestamped subtitles for any video using Gemini 2.5 Pro.

1. Open Google AI Studio

   Visit AI Studio in your browser and sign in using your Google account. Once you're logged in, you’ll be taken to the dashboard.

2. Choose the Model

   From the model-selection menu on the right, select Gemini 2.5 Pro. This version offers improved handling of structured content such as transcripts and subtitles.

3. Upload Video File

   Add the video you’d like to work on by using the upload function. Ensure the file finishes uploading before continuing.

4. Use Subtitle Prompt

   In the message field, type the following instruction:

   "Analyze this video and provide a precise transcript with timestamps (HH:MM:SS) in [specify language]."

   Replace [specify language] with your desired translation language—such as German, Japanese.

5. Review Output

   Once the subtitles are generated, review the structure. You should see clearly formatted lines with timestamps, like:

   00:02:45 - [Translated line]

6. Repeat Prompt if Needed

   Occasionally, Gemini might not provide correctly timestamped results on the first try. If that happens, simply send the same instruction again. It typically produces accurate results after a second attempt.

7. Save Subtitles

   After confirming everything looks correct, you can copy the subtitles and organize them into your desired subtitle format (such as .srt or .vtt) for use in editing tools or video platforms.

Tip: If the transcript seems off, try trimming unnecessary parts from the video before uploading—it can help Gemini stay more focused.

Top AI & Tech News

• Meta reorganized its AI division into Meta Superintelligence Labs (MSL), centered on TBD Labs led by Alexandr Wang to focus on foundation models

• Thousands of user conversations from Elon Musk’s xAI chatbot Grok that were shared via generated URLs have been indexed by search engines

• Dozens of Microsoft employees occupied Redmond campus to protest reports that Israel used Microsoft Azure to store surveillance data on Palestinians in Gaza

• Baidu's Q2 revenue fell 4% to ¥32.71B as a 15% drop in core ads outweighed a 27% rise in cloud revenue, and AI investments haven’t yet offset the ad slump

AI Art Spotlight

Model: Midjourney V7

Prompt:

Ultra-realistic 8K macro DSLR photograph of a perfectly crystal-clear water droplet suspended in mid-air. Inside the droplet is an entire raging ocean storm towering waves crashing, lightning bolts illuminating dark clouds, and a tiny, detailed wooden ship battling the sea. Razor-sharp textures of both the droplet surface and the miniature storm inside. Physically accurate refractions and reflections, high-contrast cinematic lighting, zero blur, realistic depth of field, Canon R5 + 100mm macro lens look --ar 9:16 --quality 2 --raw --v 7

Prompt of the Day

Human Mode Prompt for Clear AI Writing

This prompt instructs AI to use plain English, short sentences, and a natural, human-like conversational tone. It avoids hype, clichés, and marketing jargon. It emphasizes active voice, removes unnecessary adverbs and filler words, and requires concrete examples or steps to support claims.

Top AI Video Tutorial

8 Useful AI Tools You’ll Wish You Knew Sooner

Complimentary AI Course of the Day

LLM from Scratch – Code & Train Qwen 3 [YouTube]

A hands-on video course showing how to build, train, and understand Qwen 3 — a modern LLM by Alibaba Cloud — mixing clear conceptual explanations, implementation walkthroughs, optimization tips, and a live small-GPU training run on Google Colab.