#031 • 29 July 2025

Contents of this Newsletter

🌌

Branding with AI

⏰ : 4 min

📓

Blog Picks

⏰ : 11 min

🏣

Workforce Reset in Indian IT

⏰ : 3 min

🗞️

Weekly AI News

⏰ : 6 min

MAIN STORY

Stepping into AI-First Branding

We weren’t keen about branding when we started building Quash. We just wanted to ship. But somewhere along the way, while watching yet another "AI-powered" site pop up with the same gradients and bold sans-serifs; we knew we had to figure out how we wanted to show up.

So we asked: what would it mean to actually brand an AI-first product in a way that felt original?

Since January, we’ve been running tests in the background. We tried Midjourney, GPT-4o, Flux, Recraft. Subscribed to all of them and experimented with prompts. Broke them and refined them and tried (multiple times).

The idea wasn’t to make something "AI-generated," because that’s easy. The goal was to create an illustration style that only AI could pull off and not just hard to recreate manually, but borderline impossible without serious skill and time. Something weird, unnamed and unmistakable.

We already had our brand colors and fonts, but we wanted to build the identity further and refine it to a good extent. Illustration felt like the most impactful surface left to experiment with.

We also realized that AI can handle the style, but the subject had to come from us. That part was a conscious choice. We picked natural elements because we wanted to humanize AI to make it feel approachable, not distant or robotic. We landed on nature, skies, clouds, sceneries because it felt like the right tone. We wanted to strip AI of its robotic edge and make it feel human and relatable.

What We Settled On, and Why It Works

After hundreds of images and way too many prompts, we landed on something that felt right. The current style looks like a rough, expressive painting. Soft textures, muted gradients, lots of detail. It isn’t brand new. But it stands out in today’s world.

*Iterations that led to the final style*

More importantly, we found a way to reproduce it at scale.

Recraft became our go-to tool once we had enough visual references. It wasn’t perfect out of the box, but with a few dozen examples and a whole lot of back and forths, it started to work. We trained it to match what we needed.

None of this was accidental, but a few things did emerge organically. The orange-and-blue palette, for example, just happened. It kept showing up in the versions we liked best, and also aligned so well with our yellow brand tone.

We didn’t invent a completely new style. But we did something better suited to where we are: we created a consistent, good-looking visual identity we can own and scale without hiring a 10-person design team. The original goal to invent something from scratch is still out there. We haven’t stopped chasing it. But we’re happy with where we landed, for now.

Who Influenced Us, and Where We’re Headed

A few people deserve credit for nudging us in the right direction:

@ohneis652: probably the only person talking about AI-first branding early on, from scratch.
Jeremiah Waren, Escha, and Phi Hoang: some of the designers behind Perplexity’s visual identity (and Comet). We’d seen their work long before they were at Perplexity.

We still don’t have a name for this style and that’s fine. The goal wasn’t to name it. It was to make something that felt like us.

We’ve started using this style in blog covers and social posts. It’ll show up in more places soon. But the experimentation hasn’t stopped. We’re still messing around, trying to build something we haven’t seen yet.

Check out our Recent Blogs

industry spotlight

India’s Largest IT Layoffs Isn’t About AI

TCS is laying off 12,000 employees. It’s their biggest reduction ever and it’s not because of AI, according to CEO Krithivasan. The company says most of the exits are due to performance issues and role mismatches, primarily in overseas markets like the U.S. While hiring in India continues, a move this large from a company like TCS which is known for stability and caution; is a clear sign that the IT services model is under strain. For years, firms could get away with large benches and bloated teams. That’s no longer the case.

The layoffs will largely affect middle and senior management and are expected to play out over the 2026 financial year. TCS says this is part of a broader strategy to become “future-ready” with a focus on AI investments, tech modernization, and global market expansion. At the same time, the Ministry of Electronics and IT is reportedly reviewing the layoffs, citing concern about employment trends and the underlying signals from one of India’s largest tech employers. The ministry continues to push initiatives like the Employment Linked Incentive, which aim to boost job creation through skilling and reskilling which are areas TCS is also investing in internally to retrain and redeploy talent.

At Quash, we don’t take news like this lightly. Job cuts affect lives and while TCS is reportedly supporting affected employees through transition programs and outplacement services, it’s still a difficult moment for many. But this isn’t just about one company. It reflects a wider shift in the industry: delivery teams are now expected to run lean, stay close to the business, and prove their impact. Expectations are being reset and that opens the door for teams to rethink how they work, how they hire, and how they measure value.

news

This Week in AI

China and the U.S. Split on AI Diplomacy
Just days after Trump’s draft AI roadmap signaled an aggressive push for American dominance with heavy emphasis on deregulation, talent control, and domestic industrial power, China released its own vision at WAIC 2025. But instead of confrontation, Beijing is calling for international cooperation, open governance, and guardrails on misuse. Framing AI as a “shared global opportunity,” China is pitching itself as the diplomatic counterweight to U.S. AI nationalism. The race is no longer just about models and about who gets to write the rules.

Meta Hires ChatGPT Co-Creator as AI Chief Scientist
Meta has appointed Shengjia Zhao (one of the co-creators of OpenAI’s ChatGPT) as Chief Scientist of its new Superintelligence Lab. The hire signals Meta’s serious push into frontier AI, as it ramps up talent acquisition and multi-gigawatt infrastructure for its superintelligence ambitions. The group is led by Alexandr Wang, former CEO of Scale AI, who joins Meta as Chief AI Officer. Nat Friedman, ex-GitHub CEO, co-leads the division, overseeing applied research and AI products

Alibaba Launches Qwen‑MT: Translation for 92 Languages at Enterprise Scale
Alibaba has unveiled Qwen‑MT, a domain-aware translation model based on Qwen3 that delivers high quality across 92 languages and dialects, including low-resource ones. It rivals larger models like GPT‑4.1 and Gemini 2.5 Pro in benchmarks, while offering lower cost and faster inference at around $0.50 per million output tokens

All Newsletter Issues | Security | Contact Us | Blogs | Test Case Generator

Unsubscribe · Preferences

Meta’s New AI Chief 🚀 + Our Experiments with AI Branding ✨