🗞️In this edition

• Picsart Take Over (sponsored) - AI-powered creativity, enterprise-ready performance, and one of the best deals of the year.

• SoftBank's $5.8B Nvidia exit explained

• OpenAI weighs building consumer health products

• Google discovers AI-powered Malware that rewrites itself

• Picsart community feedback

• In other AI news –

  • Anthropic expected to reach profitability faster than OpenAI

  • Ex-Google and Meta execs raise $100M for new AI server startup

  • Meta open-sources omnilingual AI model that transcribes 1,600+ languages

  • 4 must-try AI tools

Welcome to the Picsart Takeover Edition of the OpenTools newsletter.

This week, we’re putting the spotlight on Picsart, the creative engine behind millions of visuals online — and the team closing out their biggest sale of the year.

To mark the final days of their 50% off event (up to 85% off on annual plans), Picsart is unveiling their latest AI upgrade, Remove Background v10 — faster, smarter, and built for creators who move quick.

Innovation meets deadline. And this one’s almost up.

The Remove Background v10 release takes everything teams love about Picsart: simplicity, speed, and precision — and levels it up.

Complex edges, tricky shadows, car shots, apparel photos - all handled automatically with AI you can even guide using text prompts (“keep the front car”, “add clean white background”, etc.).

It’s the same tech trusted by brands like PopSockets, Tervis, and VTMode.nl - now available for a fraction of the price.

Up to 85% off ends soon

SoftBank sells entire Nvidia stake

What's happening:

SoftBank dumped its entire Nvidia position Tuesday for $5.83 billion profit. Chairman Masayoshi Son is selling every share to fund bigger AI bets, primarily OpenAI.

The sale comes after SoftBank re-entered Nvidia in 2020 following an earlier exit in 2019. By March 2025, their stake was worth around $3 billion. Nvidia's valuation has since jumped $2 trillion, making it the first $5 trillion company in history.

Son needs liquid capital now. He's committed to Stargate, the $500 billion AI infrastructure project he announced with Trump, Sam Altman, and Larry Ellison in February. That means funding data centers and AI robot manufacturing sites across the US.

SoftBank's OpenAI stake has already gained $14.6 billion in value. Their Q2 net income hit $16.2 billion, crushing the $2.7 billion analyst expectations.

Nvidia shares dropped 1.3% in pre-market trading after the news broke.

Why this is important:

This isn't a bearish call on Nvidia. It's Son repositioning for where he thinks the real value accrues.

Chips are infrastructure. OpenAI and its ecosystem are the applications layer. SoftBank's betting the margins and defensibility sit higher up the stack.

The sale also signals Son's conviction that OpenAI's growth trajectory beats Nvidia's from here, even as questions mount about whether $1 trillion+ in AI spending by Meta, Alphabet, and others actually delivers returns.

Our personal take at OpenTools:

Son's track record is polarizing, but this move makes sense.

Nvidia ran from a $3B position to $5.83B profit on a company now worth $5T+. That leaves less upside than OpenAI's potential path from current valuation to dominance. He's trading a great company for what he thinks is a generational one.

The $5.83B profit funds Stargate and keeps SoftBank positioned at every layer: chips (Ampere Computing), infrastructure (Stargate), and models (OpenAI). Son's building vertical integration without operational headaches.

Nvidia will be fine. But OpenAI might be SoftBank's Alibaba 2.0, and Son's not missing that twice.

Read More

OpenAI considering consumer health products including AI assistant

What's happening:

OpenAI is weighing building consumer health products, including a generative AI-powered personal health assistant, as the ChatGPT maker aims to move beyond its core offerings.

OpenAI's healthcare push follows strategic hires, including Nate Gross, co-founder of physician network Doximity, as head of healthcare strategy in June, and former Instagram executive Ashley Alexander as vice president of health products in August.

At the HLTH conference in October, Gross said ChatGPT attracts about 800 million weekly active users, many seeking medical advice.

Tech giants such as Google, Amazon, and Microsoft have long tried to give consumers control over their medical data, often with limited success.

Google shut its health record service in 2011 due to low traction, while Amazon wound down its Halo fitness tracker business in 2023. Microsoft's HealthVault platform also failed to attract widespread adoption.

Why this is important:

Eight hundred million weekly users seeking medical advice on ChatGPT represents a massive existing demand OpenAI could monetize.

Hiring Doximity cofounder and former Instagram exec for healthcare roles signals serious investment, not exploration.

Healthcare is one of few consumer markets large enough to justify OpenAI's revenue projections. The company expects "hundreds of billions" in revenue by 2030. Healthcare spending in the US alone is $4.5 trillion annually.

But Google, Amazon, and Microsoft all failed at consumer health despite massive resources. That's three of the world's largest tech companies striking out.

The failures share common problems: consumers don't want tech companies controlling medical data, healthcare providers resist integration, regulatory compliance is expensive and complex.

Our personal take on it at OpenTools:

OpenAI's walking into the graveyard of failed consumer health products.

The 800 million users seeking medical advice on ChatGPT is encouraging, but that's free advice with no liability. Offering formal health products means regulatory compliance, liability insurance, and potential malpractice exposure. Totally different risk profile.

Healthcare providers resist tech integration. Electronic health records are still fragmented. Getting doctors to adopt new platforms is an expensive, slow process requiring sales teams and support infrastructure.

HIPAA compliance is a massive operational burden. Consumer health products handling protected health information need extensive security, audit trails, and breach notification processes. That's not OpenAI's core competency.

The Instagram exec hire for health products VP is interesting. Instagram's strength was consumer engagement and retention. If OpenAI's building a consumer-facing health assistant, they need someone who understands consumer products, not just healthcare.

The Doximity cofounder hire makes more sense. He built a physician network that achieved meaningful adoption. That expertise matters if OpenAI wants provider buy-in.

Still, three of the world's largest, best-resourced tech companies failed at this exact thing. OpenAI needs strong reason to believe AI changes enough to overcome the same obstacles. 

Read More

Real Results from Real Teams - Picsart Testimonial

A Family Business That Found Its Visual Edge

Sometimes the best proof of great tech isn’t a Fortune 500 name — it’s a family-run shop doing big things with small resources.

Meet VTMode.nl, a third-generation apparel retailer from the Netherlands. With one local store and an online shop, their challenge was simple but relentless: dozens of products, each with different photo styles, all needing to look consistent and professional online.

“With Picsart, we can standardize our visuals and create a consistent look and feel across our site. The presentation is cleaner, more trustworthy — and it all happens automatically.”
— VTMode.nl Team

What sealed the deal for them?
A clear, fast API. A fair price. And a platform that made it easy to turn messy product photos into a polished storefront experience.

Now it’s your turn — simplify your visuals and save big while you do it.
Get up to 85% off before the sale ends

Google discovers self-rewriting AI Malware called PROMPTFLUX

What's happening:

Google's Threat Intelligence Group found new malware called PROMPTFLUX that uses AI to rewrite its own code while attacking.

The virus uses the same kind of AI that powers Google Gemini to constantly change itself so antivirus programs and security systems can't recognize it.

Instead of following pre-written, fixed script like most malware, this thing can "learn" and "adapt" on the fly, creating new attack methods as it goes.

Google says it looks like it's still in the testing phase. Samples had unfinished parts, and it hasn't been seen in the wild infecting anyone. Google shut down assets and accounts tied to it.

Why this is important:

This is one of the very first real-world examples of malware using AI to evolve on its own.

Traditional antivirus works by recognizing the "signature" of a virus, a digital fingerprint. But if the virus can change its own fingerprint every few seconds, detection becomes nearly impossible.

This kind of adaptive malware could make old virus detection methods almost useless, forcing the entire security industry to find new strategies.

Google's worried a "black market" for these kinds of AI tools will pop up, meaning even low-skill amateur hackers could get their hands on super-advanced, shape-shifting malware.

Google is rolling out a new security framework specifically to defend its AI systems. They're building "counter-AI" programs (one is literally named "Big Sleep") designed to hunt down and patch vulnerabilities before AI-powered threats can use them.

Our personal take on it at OpenTools:

This is the cybersecurity nightmare everyone predicted.

AI malware that rewrites itself on the fly breaks signature-based detection fundamentally. If the virus changes its code constantly, traditional antivirus can't keep up. That's most consumer and enterprise security tools obsolete.

The good news is it's still in testing and hasn't infected anyone. The bad news is someone built it, which means others will too. This genie doesn't go back in the bottle.

"Black market for evil AI tools" is the key concern. Right now, building sophisticated malware requires expertise. AI-powered malware-as-a-service democratizes advanced attacks. Amateur hackers with AI malware kits become as dangerous as expert attackers.

Google building "Big Sleep" and other counter-AI programs shows they're taking this seriously. But it's now AI versus AI, and attackers have an advantage. Defense needs to catch every attack. Offense only needs to succeed once.

The "whose AI is smarter" framing is accurate but terrifying. Security becoming an AI arms race means constant escalation. Attackers improve AI, defenders improve AI, attackers improve again. No stable equilibrium.

This validates concerns about dual-use AI technology. Same models powering helpful assistants can power adaptive malware. Can't prevent one without limiting the other.

Read More

📢 What else is happening?

• Anthropic Is on Track to Turn a Profit Much Faster Than OpenAI – Financial documents from both companies show the different approaches they are taking to the AI boom.

• Former Google, Meta executives raise $100 million for high-capacity AI servers startup – Three former Meta and Google silicon executives on Monday announced they’ve raised $100 million to build technology they say will reduce cloud companies’ spend on data center buildouts.

• Meta returns to open source AI with Omnilingual ASR models that can transcribe 1,600+ languages natively – It’s a shift from static model capabilities to a flexible framework that communities can adapt themselves. 

👩🏼‍🚒Discover mind-blowing AI tools

1. Picsart API — Trusted by teams like Deutsche Telekom and FujiFilm.
   The Remove Background v10 update is live — faster, smarter, and now up to 85% off.

2. Taranify - A platform that uses AI technology to provide mood-based recommendations for music

3. CodeWiz - An AI-powered tool designed to assist developers in their coding journey

4. Sloped - Helps users turn their data queries into shareable and easy-to-read dashboards