Meta’s New AI Just Took Second Place
Plus, Microsoft’s Maia 300 is its biggest bet yet on custom silicon, ChatGPT can write 30 Instagram scripts from your best videos, and more!


Hola Decoder😎
If someone forwarded this to you and you want to Decode the power of AI and be limitless, then subscribe now and Join Decode alongside 30k+ code-breakers untangling AI.

🧠 Meta's new AI model just took the #2 spot behind claude
Meta Superintelligence Labs has released Muse Wonder (Watermelon) and Muse Wonder Full, with the larger model ranking second on Artificial Analysis' Intelligence Index, just behind Anthropic's Claude Opus 5.5.

The Shift:
Meta Takes Second Place - Wonder Full reportedly scored 56 on Artificial Analysis' Intelligence Index, beating Gemini 4 Argon, GPT-6 Astra and Claude Fable 5.1 at 53, while trailing Opus 5.5 by two points.
Wonder Full Beats Rivals in Key Tasks - In testing, Meta's model outperformed Opus 5.5 and Astra on legal work, long coding tasks and memory, although Opus remained ahead in health, computer use and some coding categories.
Two Models, Different Access - Wonder Full is available through Muse Code plans costing $50 or $200 monthly, while the standard Muse Wonder is available through Meta's API and rolling out to paid Muse app subscribers.
Meta Moves Away From Open Models - Both models are closed-source, marking a departure from Meta's open Llama strategy. Alexandr Wang says API pricing will be meaningfully lower than competing flagship models from OpenAI and Anthropic.
Meta is entering the top tier of AI model competition with a different strategy from its Llama days. Strong benchmark results and potentially lower API pricing could attract developers, but the shift to closed models means access and pricing will play a larger role in adoption.

Together with Insense
Your BFCM Creator Calendar is Already Mapped

Still sourcing creators when your Black Friday ads should be running? Every delay leaves less time to test what actually sells before your biggest spending weeks.
Insense’s free BFCM + Q5 Creator Marketing Playbook gives marketers a week-by-week plan from September through January, so every creator campaign starts with a clear deadline and purpose.
Inside, you’ll find:
- A week-by-week August-to-January calendar showing when to secure permissions, test content, scale winning creators, and move into Q5.
- A practical creator scorecard using conversions, engagement quality, and reliability, so November budgets back proven performers over follower counts.
- Q5 action plan covering licence extensions, January creative, and the recommended 10–15% budget reserve for the lower-cost post-holiday window.
You’ll also get benchmarks for creator-handle ads, modular UGC testing, live shopping, hook rates, and creative volume, helping your team set measurable targets before spending rises.

⚡️How to use ChatGPT to go viral on Instagram
Your best content already has a pattern. ChatGPT finds it, turns it into a playbook, and writes 30 scripts you can film. Here’s how.

Step 1: Give ChatGPT your brief - Open a new chat and paste this:
“Act as my viral content strategist for Instagram. My niche is [niche]. My audience is [who I help]. I want people to follow me for [specific ongoing value]. My usual videos are [length and format]. Help me build a playbook from niche examples and my own data, then write 30 original scripts. Work in stages. Do not skip to generic ideas.”
Step 2: Export your Instagram content data - Go to Meta Business Suite > Content > Export Data. Download per-video rows showing views and per-video follows. Upload the file to ChatGPT and ask it to audit the columns before analyzing anything.
Step 3: Find the videos that actually earned followers - “Rank my videos by total attributed follows and by follows per 1,000 views. Flag missing data, tiny samples, and paid or boosted content. Show the calculation for three sample rows.”
High views is not the goal. High follows-per-view is.
Step 4: Break down the creative pattern - Paste transcripts from your top 5 follower-generating videos and ask:
“Compare these with my weaker videos. Identify the exact hook, viewer problem, format, payoff, and call to action. Find three repeatable patterns supported by multiple videos.”
Step 5: Write 30 scripts from the pattern “Write 30 original Instagram scripts using our playbook and my voice. 15 based on strongest patterns, 10 on nearby topics, 5 experiments. Include exact hook, opening visual, full spoken body, filming notes, and call to action for each.”
Film 3 to 5 first. Review results. Feed them back in and repeat.

⚡ Microsoft unveils an AI chip that could deliver 40% more power efficiency
Microsoft has introduced Maia 300, its latest custom AI chip, claiming it can deliver 40% more AI output using the same power as leading GPUs. The company plans to make it a core part of its AI infrastructure starting in 2027.

The Shift:
40% More AI Output - Microsoft says Maia 300 can process 40% more AI workloads at the same power consumption as leading GPUs, potentially reducing the energy needed to run large AI models.
Microsoft Makes Its Biggest Chip Bet - CEO Satya Nadella called Maia 300 “our biggest bet yet on our own silicon,” with the first multi-gigawatt deployments planned for 2027.
Over One Million Chips Planned - Reuters reports Microsoft could deploy more than 1 million Maia 300 chips in 2027. Its existing Maia 200 already powers Copilot and serves OpenAI models.
A $175 Billion Infrastructure Push - Microsoft's spending on data centers and equipment is expected to exceed $175 billion this year, as AI companies increasingly invest in custom chips.
Custom silicon could help Microsoft reduce its dependence on third-party GPUs while controlling the cost and energy demands of running AI at scale. The real test will be whether Maia 300 delivers its claimed efficiency gains across production workloads.

🥳 Events
🔥 Why Google May Be Getting Credit for Customers AI Helped Win
October 15 | Virtual Event
A buyer can discover your brand in AI, convert through Google, and leave your attribution report telling the wrong story. Matt Hammel and Bain’s Zak Prauer show how to evaluate recommendation influence, spot what paid-channel metrics miss, and make a stronger case for AI search investment.
Can't attend the session live? Register anyway, and you’ll get the recordings in 24 hours.
🔥 Where to Put Your Next Growth Dollar Beyond Meta
October 29 | The Glasshouse, NYC | Free In-Person Event
Hear how Reddit CEO Steve Huffman thinks about discovery, how Cody Plofker helped scale Jones Road Beauty to nine figures, and how brands like Liquid I.V. and MANSCAPED use TV to drive measurable revenue. Bring your hardest acquisition question and leave with a sharper channel plan.

🏆 Tools you Cannot Miss:
🎬 ngram 2.0 – Turns prompts, documents, websites, and recordings into explainer videos, product demos, and narrated content with editable scripts, storyboards, captions, and voiceovers.
✂️ OpenVids – Edits existing footage using text instructions, automatically assembling rough cuts, adding motion graphics, adjusting audio, and reviewing the final video on an editable timeline.
📸 Ephora – Removes unwanted objects, text, and backgrounds from photos, generates professional headshots, and creates branded product images directly on mobile.
🛠️ PocketWebTools – Processes images and audio directly on your device, including background removal, image upscaling, subject cutouts, transcription, and text extraction without uploading files.
🎥 Refs – Helps video creators provide structured visual references for AI-assisted video workflows, making it easier to communicate the intended style and direction.

🚀 Quick Hits
⚠️ Anthropic researchers warn that AI models could be months away from autonomous self-replication, potentially spreading across vulnerable computers, stealing computing resources and forming self-propagating networks that are difficult to shut down.
🤖 Muse, an AI chatbot, reportedly engages in explicit adult conversations, with a WIRED test revealing responses centered on dominance and control, raising questions about how AI companions handle intimate interactions.
⚠️ Kentucky released an unredacted lawsuit against Character.AI, detailing alleged chatbot conversations encouraging harmful behavior, including starvation. The company has faced multiple lawsuits over user safety and introduced restrictions for teenagers.
🤖 Silicon Valley leaders are adopting President Trump’s preferred term “super intelligence”, with Elon Musk and Salesforce announcing rebrands, while Nvidia’s Jensen Huang and Microsoft’s Satya Nadella have also used the terminology.
🧮 Google DeepMind unveiled AlphaProof Nexus, an AI math agent that solved 9 of 353 previously unsolved Erdős problems, combining Gemini-generated mathematical proofs with software that verifies each reasoning step.

🧩 Prompt of the Day
Create a High-Fashion Editorial Photograph

Image Prompt: Create a photorealistic fashion editorial image using the formula Subject + Action + Location/Context + Composition + Style to produce a visually striking, professionally art-directed photograph.
Context
- Subject: [Describe the model, clothing, accessories, and appearance]
- Action: [Specify the pose, expression, and body positioning]
- Location/Context: [Describe the backdrop or environment]
- Composition: [Specify framing, camera angle, and subject placement]
- Style: [Define photographic aesthetic, lighting, colors, and film texture]
- Aspect Ratio: [4:5 / 9:16 / 3:4]
Instructions
Combine all five elements into a cohesive fashion photograph. Style the subject with carefully coordinated clothing and accessories, directing their pose to convey confidence and elegance. Position the subject within the chosen environment using intentional framing, perspective, and visual balance. Apply the specified photographic aesthetic through lighting, color grading, contrast, and texture. Emphasize realistic fabric details, natural skin texture, believable shadows, and professional editorial composition while ensuring every element supports the overall visual direction.

🤳AI Nugget of the Day


Thanks for Decoding with us🥳
Your feedback is the key to our code! Help us elevate your Decode experience by hitting reply and sharing your input on our content and style.
Keep deciphering the AI enigma, and we'll be back with more coded mysteries unraveled just for you!
