#024 • 10 June 2025

Contents of this Newsletter

🎥

AI in Other Industries

⏰ : 6 min

🤺

Reddit Vs Anthropic

⏰ : 9 min

👩🏻‍🏫

New Team Members

⏰ : 4 min

🗞️

Weekly AI News

⏰ : 6 min

MAIN STORY

AI’s Doing Everything But Who’s Testing It?

Beyond code: a look at how AI is quietly rewriting other industries

Everyone talks about AI in software right now. We get that because we do too. From code generation to mobile testing and everything in between, AI is changing how teams build software. But let’s be honest: it’s also gotten a little one-track. There’s more to this wave than just better IDE plugins.

So this week, we wanted to zoom out. Because while the dev world talks prompts and pipelines, other industries are quietly reinventing themselves with AI, and some of the stuff they’re doing is wild.

Here’s a look at some of the more interesting ways AI is showing up in other industries:

Filmmaking

Google’s Veo 3 is rolling out globally, enabling entire cinematic scenes, like sweeping drone shots, CGI backdrops, and even character animations that are generated in minutes instead of weeks. What once cost millions is now a prompt away. There’s also a growing discourse among filmmakers and studios about how to maintain creative coherence and authorship in this new paradigm.

Agriculture

AI is now at the heart of modern farming. Tools like Blue River Technology use computer vision and machine learning to monitor crop health, detect pests, and optimize planting cycles. It’s helping farmers make smarter, more sustainable decisions, with a level of precision never seen before.

Education

AI is making serious inroads into classrooms. Platforms like EDMO act as intelligent copilots for students, offering adaptive, real-time guidance that helps personalize the learning journey. As it takes root in foundational spaces like education, the opportunity is massive—but it also means we need to tread thoughtfully, ensuring that what gets taught is accurate and equitable.

Law

LucioAI is transforming legal workflows by automating contract reviews, organizing case files, and accelerating legal research. It's allowing lawyers to focus more on strategy and nuanced interpretation. At the same time, incidents like the recent UK court case where AI-generated, fabricated precedents were mistakenly submitted—remind us why checks and balances matter.

Sustainability

Companies like Mankomb are using AI into everyday appliances to improve waste management and energy efficiency. From converting wet waste into soil enhancers to managing home microgrids, these innovations are pushing us toward greener living—while also making critical systems more dependent on AI-driven decisions.

Blockchain

Tools like ChainGPT are making blockchain more secure by automating smart contract audits and monitoring DeFi transactions for anomalies. In a space where a tiny error can cost millions, the ability to scale vigilance with AI is a game-changer.

So What Happens When AI Gets It Wrong?

As AI starts touching deeper systems—legal, medical, environmental, educational, the stakes get higher. Bugs and biases in these contexts aren’t just technical flaws; they’re systemic risks. For example, misdiagnoses from flawed AI-generated radiology summaries or AI-driven legal tools fabricating case law have already caused real-world harm. These are not edge cases—they’re reminders of why AI quality can’t be an afterthought. That’s why we believe testing AI isn’t just a development need. It’s a societal one.

That belief has guided our work at Quash. Over the past months, we’ve been heads-down, working with mobile teams and getting thoughtful feedback from teams already putting Quash through its paces. And it’s made us more confident than ever about what we’re building.

Now, we’re almost ready to open the doors. Stay tuned because something big is on its way.

Check out our Recent Blogs

social

Reddit Sues Anthropic Or Should We Say, Sam Altman Sues Anthropic

The image describes the clash between Reddit and Anthropic, with Sam Altman having a major stake at Reddit

Reddit has officially sued Anthropic for allegedly scraping its platform to train AI models without permission. In the complaint, Reddit claims Anthropic violated its terms of service and used the content for commercial gain, without any licensing deal in place. What makes this especially interesting is that Reddit has already signed licensing deals with OpenAI and Google, allowing them to use Reddit’s data under certain terms that, it says, protect user privacy and interests. Anthropic, on the other hand, reportedly declined to engage when Reddit approached them.

And here’s where it gets interesting: Sam Altman, CEO of OpenAI and a direct competitor to Anthropic, holds an 8.7% stake in Reddit, making him the third-largest shareholder. He also used to sit on Reddit’s board. So while Reddit is calling out one AI startup for unauthorized data use, it’s simultaneously monetizing that same data through another AI company led by one of its largest stakeholders. Reddit’s Chief Legal Officer didn’t hold back, stating they won’t tolerate “profit-seeking entities like Anthropic commercially exploiting Reddit content for billions of dollars without any return for redditors or respect for their privacy.” This case stands out not just for the irony, but because it’s the first time a major tech platform (not a publisher or artist, as we’ve mostly seen so far) has taken an AI company to court over training data.

Welcome To The Team!

This week we’re happy to welcome 3 new members to our team: Ayushi, Uzma and Omkar.

Omkar

Fun Fact: I have five cats and I spend most weekends deep in a TV show, currently diagnosis fictional illnesses with House M.D.

Hi, I’m Omkar and I’m a PGDM student at JAGSoM with a background in Computer Science. I’ve built AI tools and user models that blend tech with strategy. I am excited to join Quash as a product intern.

Uzma

Fun Fact: I’m a certified bathroom singer with a soft spot for music, K-dramas, and any movie that involves zombies or monsters.

Hi, I’m Uzma, and I’ve joined Quash as a design intern. I love solving problems through design—whether it’s UI/UX, branding, or motion. I’m all about the details and making things work, not just look good.

Ayushi

Fun Fact: I'm a selectively extroverted, gym-obsessed dog magnet who firmly believes that veg biryani is a scam.

Hi, I’m Ayushi, joining Quash as a product intern. I’m a Mechanical Engineering grad from VNIT Nagpur and have built everything from automated production analysis to scalable cloud security tools.

news

This Week in AI

WWDC 2025: A Quiet Year for Apple Intelligence + “Liquid Glass” Backlash
Apple Intelligence expands across devices with on-device Genmoji, live translation, and ChatGPT-powered image tools. A new Foundation Models Framework lets devs build privacy-first AI features right into their apps. A smarter Shortcuts app taps into Apple’s LLM (and optionally ChatGPT), bringing intelligent automation with minimal code. But there’s no Siri overhaul or AI “wow” moment, Apple is clearly treating 2025 as a setup year, focusing on under-the-hood upgrades and developer tooling. The new Liquid Glass UI debuts across platforms, though early reactions from users are mixed.

Meta Moves to Replace Human Risk Reviewers with AI
Meta plans to automate 90% of its privacy and integrity reviews using AI, per internal docs obtained by NPR. Previously human-led, these assessments cover risks from algorithm changes to content moderation. AI will now handle even higher-stakes areas like youth safety and misinformation, with engineers gaining more decision-making power via automated risk tools.

MemVid: Lightning-Fast Semantic Search, No Database Needed
MemVid is a new open-source tool that turns millions of text chunks into a single MP4 file enabling sub-second semantic search, high-speed retrieval, and up to 10x compression over traditional vector databases.Built for efficiency, MemVid supports PDF import, offline-first use, plug-and-play LLMs, and a simple Python API.

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

AI’s Doing Everything—But Who’s Testing It?