Instagram | Sponsor this newsletter | Let’s Connect!

My fellow AI explorers

GPT-5.1 is now the default for millions without a flashy benchmark chart, image models can spin a single shot into a full photo set, and the tools in the background, transcription, deep research, coding platforms, are leveling up in ways that make everything feel smoother and more “alive.”

In today’s edition:

• 🤖 GPT-5.1: A denser, more human default model

• 📸 One photo → full carousel with Qwen Image Edit Angles

• 🎧 11Labs Scribe V2 and the transcription arms race

Must See AI Tools

• 💰 Payman: AI That Pays Humans. Over 10,000+ signed up for the beta

• 💫 SubMagic: An AI tool that edits short-form content for you! (Get 10% off using code “uncoverai” at checkout)

• 🎤 11Labs: #1 AI voice generator (Click Here to get 10,000 free credits upon signing up!)

• 🤖 ManyChat: Automate your responses & conversations on IG, FB and more! (Click Here to get first month for free)

• 🎙️ Syllaby: The only social media marketing tool you’ll ever need - powered by AI! (Get 25% off the first month or any annual plan with code “UNCOVER” at checkout)

Find your customers on Roku this Black Friday

As with any digital ad campaign, the important thing is to reach streaming audiences who will convert. To that end, Roku’s self-service Ads Manager stands ready with powerful segmentation and targeting options. After all, you know your customers, and we know our streaming audience.

Worried it’s too late to spin up new Black Friday creative? With Roku Ads Manager, you can easily import and augment existing creative assets from your social channels. We also have AI-assisted upscaling, so every ad is primed for CTV.

Once you’ve done this, then you can easily set up A/B tests to flight different creative variants and Black Friday offers. If you’re a Shopify brand, you can even run shoppable ads directly on-screen so viewers can purchase with just a click of their Roku remote.

Bonus: we’re gifting you $5K in ad credits when you spend your first $5K on Roku Ads Manager. Just sign up and use code GET5K. Terms apply.

Use code GET5K now

Open AI

GPT-5.1: Denser, Calmer, Weirdly Human

OpenAI just rolled GPT-5.1 out as the new default model—and instead of a victory lap of benchmarks, they basically said: “Here, try it.”

Early tests point to something you feel more than you can chart: 5.1 is denser, calmer, and more intentional in how it writes and formats. Less “eager intern,” more “person with life experience who has learned to pick their words carefully.”

What stands out so far:

• Tone upgrade: Feels more human, less sugary, and more willing to suggest solutions instead of just agreeing.

• Better defaults: Emails and ideation prompts come out concise by default—more like Claude, less like old GPT walls of text.

• Formatting & instruction-following: Clean markdown, headings, quotes, and variable placeholders make it easier to skim and reuse.

Put GPT-5, 5.1, and Claude side by side and a pattern emerges:

• For simple tasks (like “email my boss about the broken coffee machine”), 5.1 keeps things short, polite, and concrete.

• It leaves unknown details as variables (like names, dates, or specifics), which is now becoming the standard for good models—no more hallucinating fake details for your real life.

• On deeper, personal or reflective prompts, 5.1’s responses feel unusually dense—less fluff, more “this hurt because…” energy. It’s harder to skim precisely because more of the text actually says something.

On ideation (YouTube thumbnails, titles, hooks):

• 5.1’s ideas lean heavily into curiosity and clear emotional cues (“It’s getting too smart…”; “This tool will save you hours”) without going full unhinged clickbait.

• Claude is still king of chaotic, viral thumbnail ideas—but 5.1 has clearly moved up into “actually useful for creators” territory and formats everything in a way that’s ready to copy-paste.

🔮 Prediction: GPT-5.1 becomes the “everyday driver” for millions, not because it’s radically smarter at math or code—but because it feels more like a co-writer with taste. Over the next year, expect a wave of products to quietly reorient around this “denser but still concise” style: dashboards with more structured insights, agents that respect your time, and creative tools that give you fewer—but better—options by default.

Try it out here

Founders…

🤖 Need an AI Agency to Help Your Business Implement AI Solutions?

Our preferred partner Align AI provides you with an expert AI and Automation implementation team to add 10-40+ hours of increased productivity per employee and achieve your goals faster.

Click here to schedule your company’s complimentary AI and Automation Strategy Plan

Qwen Image Edit Angles

One Photo, Many Angles: Qwen Turns Still Shots into Carousels

Imagine taking one good photo from your wedding, a product shoot, or a stage talk… and spinning it into an entire photo album with new angles that never existed.

That’s the promise behind Qwen Image Edit Angles—a model tuned specifically to generate fresh viewpoints of a single source image. From a capybara to a keynote speaker shot, you feed it one angle and it reconstructs plausible side views, alternate perspectives, and new compositions.

Key details:

• Single input, multiple views: Drop in an image and ask for “side view,” “three-quarter angle,” or other directions—the model hallucinate-consistent new angles.

• Surprisingly coherent context: Clothing, background, and lighting hold up well; structure of the scene is preserved even as viewpoint shifts.

• Faces are “good enough,” not perfect: It’s strong on pose and clothing continuity, but still a bit off on fine facial identity in some outputs.

Why this matters more than it first appears:

• For creators, one good hero shot can now become a full Instagram carousel: stage-left, stage-right, close-ups, faux-BTS—all generated from the same original.

• For brands and e-commerce, you can prototype multi-angle product photos without a studio session… then selectively reshoot only the best-performing views.

• For events (weddings, conferences, summits), you get “alternate angles” that never existed—enough to tell a full visual story when the original photographer only captured a handful of shots.

There are practical caveats: inference isn’t free (Replicate runs cost cents per image), popular free demos get jammed, and you still need to sanity-check identity and small details. But directionally?

🌊 Prediction: Multi-angle generation from single uploads will become table stakes for creative tools in 2025–26. Instead of “upload 10 photos,” you’ll upload one great one—and your editor will propose a carousel, a short video pan, and a mini lookbook automatically. Image models are quietly turning into scene models.

Try it out for free here

AI SaaS Founders

🚨Want Millions of Impressions For Your AI SaaS, Done For You?

At uncovernews.co, we specialize in getting AI SaaS products the attention they deserve through strategic influencer marketing campaigns designed to drive millons of impressions at the fraction of the cost!

Get Your AI Startup’s News or Product In Front of Millions Quickly

11Labs Scribe V2

11Labs Scribe V2: Transcription That Outruns Your Patience

11Labs dropped Scribe V2, and by their benchmarks (and plenty of real-world testing), it’s now the transcription model to beat. We were already in a good place with Whisper-style models—this pushes the bar further.

What’s new & notable:

• State-of-the-art accuracy: Outperforms previous leaders on standard benchmarks and in messy, real audio.

• 90+ languages: Serious coverage, including non-English content where many tools still fall apart.

• Fast: ~150 ms latency: Fast enough for near-real-time feedback in interactive tools.

If you’ve been using Scribe V1 or Whisper-based tools, the upgrade feels incremental but important:

• Noisy environments (conferences, summits, webinars) are more forgiving—Scribe V2 can handle audio that most humans would give up on.

• It’s already being wired into multi-step pipelines: record → transcribe → summarize → chapterize → generate clips and titles.

• For podcasters, YouTubers, and educators, you’re getting closer to “upload once, get an entire content stack back” without manual cleanup.

📈 Prediction: Over the next 12–18 months, transcription will stop being a “feature” and become invisible infrastructure. Every serious product in meetings, education, content, or customer support will quietly run SOTA transcription under the hood. The real differentiation shifts to what happens after the transcript: atomic highlights, auto-courses, synthetic hosts, and personalized “replays” tailored to how you learn.

Other Relevant AI News!

🧠 Kimi K2, a new open-source “thinking” model from Kimi, is getting strong early reviews for deep reasoning and coding—especially in Chinese—earning praise in this hands-on review.

🎥 Google Flow’s latest update brings richer camera path and angle controls to its video animation studio, making AI video feel more like directing a real camera—see the official update on X.

🛠️ Replit’s new AI integrations let you wire external APIs and tools into your Replit apps without juggling a dozen API keys, pushing “vibe coding” even closer to full-stack app building—details in their launch thread.

🌍 Meta’s omni-lingual ASR is targeting ~500 low-resource languages, expanding speech recognition far beyond English and a handful of major tongues—this kind of work is increasingly highlighted in broader State of AI reports.

🧩 Sakana’s Sudoku-GPT5 experiment shows how structured, multi-step reasoning pipelines can beat brute-force LLM prompting on logic puzzles—read their explainer to glimpse where “tool-using agents” are heading.

🧾 OpenAI vs. The New York Times continues: NYT reportedly requested access to over 20M ChatGPT conversations to probe paywall abuse, and OpenAI pushed back publicly, framing it as a defense of user privacy—see OpenAI’s response here.

Golden Nuggets

• 🤖 GPT-5.1 doesn’t shout with benchmarks—it quietly becomes the new default by feeling more human, denser, and more intentional in how it writes and formats.

• 📸 Qwen Image Edit Angles turns single photos into multi-angle scenes, hinting at a near-future where carousels, lookbooks, and albums start from one good shot, not thirty.

• 🎧 Scribe V2, Meta’s ASR, and Kimi K2 show how fast the “infrastructure layer” is moving—transcription, low-resource languages, and open-source reasoning are all catching up to the headline models.

Would love to hear your thoughts on GPT 5.1? Send me your thoughts by replying to this email (yes, I read them all :)

Until our next AI rendezvous,

Anthony | Founder of Uncover AI