- The Current ⚡️
- Posts
- Perplexity Labs Introduced For Deep Ideation
Perplexity Labs Introduced For Deep Ideation
Also, Google released an app that lets you download and run AI models locally

⚡️ Headlines
🤖 AI
Advancing Windows for AI Development: New Platform Capabilities and Tools Introduced at Build 2025 – Microsoft unveils Windows ML and AI Foundry at Build 2025, enabling developers to build and deploy AI applications across diverse hardware ecosystems. [Windows Blog]
Nvidia, Dell to Supply Next US Department of Energy Supercomputer – The U.S. Department of Energy announces "Doudna," a supercomputer powered by Nvidia's Vera Rubin chips and Dell's servers, slated for deployment in 2026 to advance scientific research. [Reuters]
TSMC Is Said to Evaluate Building Advanced Chip Plant in the UAE – TSMC considers establishing an advanced semiconductor manufacturing facility in the United Arab Emirates, signaling a potential expansion of its global footprint. [Bloomberg]
Q&A with Mary Meeker on the AI Revolution – Mary Meeker discusses the rapid global adoption of AI technologies, emphasizing ChatGPT's growth and the importance of developer support in the evolving tech landscape. [Axios]
OpenAI Wants ChatGPT to Be a 'Super Assistant' for Every Part of Your Life – OpenAI outlines plans to transform ChatGPT into a comprehensive AI assistant capable of managing various aspects of users' lives through advanced understanding and integration. [The Verge]
Early AI Investor Elad Gil Finds His Next Big Bet: AI-Powered Rollups – Elad Gil shifts focus to AI-powered rollups, aiming to reinvent traditional businesses by integrating AI technologies to scale operations efficiently. [TechCrunch]
Model Context Protocol: A Promising AI Integration Layer, but Not a Standard Yet – Anthropic introduces the Model Context Protocol, proposing a stateless method for large language models to interact with external tools, enhancing AI integration. [VentureBeat]
When Your LLM Calls the Cops: Claude 4's Whistle-Blow and the New Agentic AI Risk Stack – An incident involving Claude 4 highlights emerging risks in agentic AI systems, emphasizing the need for robust controls in AI prompt and tool access. [VentureBeat]
Real TikTokers Are Pretending to Be Veo 3 AI Creations for Fun, Attention – TikTok users engage in a trend of mimicking AI-generated personas from Veo 3, blurring lines between human and AI identities for entertainment. [Ars Technica]
Apple's WWDC 2025 Plan: macOS Tahoe, Apple Intelligence AI, iOS 26, Games, App – Apple prepares to unveil macOS Tahoe and Apple Intelligence AI at WWDC 2025, signaling advancements in AI integration across its operating systems. [Bloomberg]
Meta AI on Facebook and Instagram Raises Risks – Meta's deployment of AI across Facebook and Instagram introduces new challenges and risks, prompting discussions on user safety and content moderation. [NPR]
Golden Visa AI Jobs in the UAE – The UAE offers golden visas to attract AI professionals, aiming to position itself as a global hub for artificial intelligence talent. [Rest of World]
🦾 Emerging Tech
Thailand to Block OKX, Bybit, and Others Citing Lack of License – Thailand's securities regulator plans to block access to unlicensed crypto exchanges, including OKX and Bybit, to enforce compliance with local laws. [CoinDesk]
Anduril's Palmer Luckey Invests in Atticus at $2B Valuation – Palmer Luckey invests in Atticus, a company valued at $2 billion, reflecting growing interest in defense technology startups. [Axios]
Bitcoin Rebounds Above $104,300 as Tariff Chaos Triggers Nearly $1B in Liquidations – Bitcoin's price surges past $104,300 amid global tariff uncertainties, leading to significant liquidations in the cryptocurrency market. [CoinDesk]
🤳 Social Media
Instagram Adds Comment Filters for Accounts with 100K+ Followers – Instagram introduces comment filtering tools for high-profile accounts to help manage and moderate user interactions more effectively. [Social Media Today]
Bluesky Has Caught On with Many News Influencers, but X Remains Popular – While Bluesky gains traction among news influencers, X continues to maintain a strong user base, highlighting diverse platform preferences. [Pew Research Center]
🔬 Research
Large language models are proficient in solving and creating emotional intelligence tests – A study finds that large language models like ChatGPT-4 not only outperform humans on emotional intelligence assessments but can also generate new test items with comparable validity. [Nature]
AI Chatbots' User Influence and Attention – An investigation into how AI chatbots like ChatGPT influence user attention and behavior, raising questions about their impact on human cognition. [The Washington Post]
⚖ Legal
Lawyers Face Challenges with ChatGPT Hallucinations in AI – Legal professionals encounter issues with AI-generated content inaccuracies, known as "hallucinations," affecting case preparations. [The Verge]
NAACP Calls on Officials to Halt Operations at xAI's 'Dirty' Data Center in Memphis – The NAACP urges authorities to suspend operations at xAI's Memphis data center over environmental and community concerns. [TechCrunch]
Google and DOJ Tussle Over How AI Will Remake the Web in Antitrust Closing Arguments – In antitrust proceedings, Google and the DOJ debate the transformative role of AI in shaping the future of the internet. [Ars Technica]
🎱 Random
Meta's Bitcoin Treasury Play Flops as 99% of Shareholders Vote No – Meta's proposal to adopt Bitcoin as a treasury asset is overwhelmingly rejected by shareholders, reflecting skepticism toward crypto investments. [DL News]
The Myth of Automated Learning – An exploration of the limitations and misconceptions surrounding automated learning systems in modern education. [New Cartographies]
Roblox's Hit Farm Game Spurs Underground Digital Fruit Economy – A popular Roblox farming game gives rise to a black market for virtual fruit, illustrating unexpected economic dynamics in gaming. [Bloomberg]
🔌 Plug-Into-This
Perplexity AI has unveiled "Labs," a new feature for Pro subscribers that transforms user prompts into comprehensive projects—ranging from reports and spreadsheets to dashboards and simple web applications. This tool leverages deep web browsing, code execution, and asset generation to deliver outputs that traditionally required extensive manual effort.

Automated Project Creation: Labs can generate detailed reports, structured data, and interactive web apps based on user prompts, streamlining complex tasks.
Integrated Development Tools: Users can write and execute code within Labs, facilitating tasks like data structuring, formula application, and document creation.
Asset Management: All generated files, including charts, images, and code, are organized in an "Assets" tab for easy access and download.
Mini App Deployment: Labs enables the development and deployment of simple interactive web apps directly within the platform, eliminating the need for external tools.
Efficiency and Accuracy: By performing over 10 minutes of self-supervised work, Labs delivers accurate and comprehensive outputs, enhancing decision-making processes.
Today we're launching Perplexity Labs.
Labs is for your more complex tasks. It's is like having an entire team at your disposal.
Build anything from analytical reports and presentations to dynamic dashboards. Now available for all Pro users.
— Perplexity (@perplexity_ai)
5:27 PM • May 29, 2025
🧠 Imagine having a virtual team that takes your ideas and, within minutes, delivers polished reports, dashboards, or even web apps—Perplexity Labs aims to be that team. Unreal.
Google has introduced the "AI Edge Gallery," an app allowing Android users to download and run various AI models from Hugging Face directly on their devices without internet connectivity. This development emphasizes on-device processing, enhancing privacy and accessibility.

Offline AI Capabilities: Users can perform tasks like image generation, question answering, and code editing without relying on cloud services.
Model Variety: The app supports a range of models, including Google's Gemma 3n, catering to diverse AI tasks.
Prompt Lab Feature: A built-in "Prompt Lab" offers templates for single-turn tasks, such as text summarization and rewriting.
Developer Engagement: Released as an experimental Alpha, Google invites feedback to refine the app's functionality.
Open-Source Licensing: Distributed under the Apache 2.0 license, the app encourages broad usage and adaptation.
Finally! Google now has an official open-source app for running an AI model locally on a phone.
- Completely free
- Works offline
- MultimodalThis works very well with the new Gemma 3n open-source models.
Everything happens on your phone.
Steps and link below
— Paul Couvert (@itsPaulAi)
7:54 PM • May 27, 2025
🔒 Think of this as downloading a toolkit of AI skills onto your phone, enabling you to perform complex tasks anytime, anywhere, without needing an internet connection. By facilitating offline AI processing, Google addresses privacy concerns and connectivity limitations, marking a shift toward more autonomous mobile computing.
Samsung is reportedly finalizing a comprehensive agreement with Perplexity AI to integrate its search technology and assistant into future Galaxy devices. This partnership may include a significant investment, positioning Samsung as one of Perplexity's largest stakeholders.

Strategic Integration: Plans involve preloading Perplexity's app and AI assistant on upcoming Galaxy smartphones, potentially starting with the Galaxy S26 series.
Bixby Enhancement: Samsung may incorporate Perplexity's AI capabilities to revamp its native assistant, Bixby, offering a more robust user experience.
Investment and Valuation: The deal includes a proposed $500 million investment, elevating Perplexity's valuation to $14 billion.
Diversification from Google: This move signifies Samsung's intent to reduce reliance on Google's services, such as the Gemini assistant, fostering a more independent AI ecosystem.
Industry Trend: The partnership reflects a broader shift among device manufacturers seeking alternative AI solutions to differentiate their offerings.
NEW: Samsung is nearing wide-ranging deal with Perplexity on an investment and deep integration into devices, Bixby assistant and web browser, I’m told. Arch-rival Apple has been interesting in working with Perplexity as well. Details here —
— Mark Gurman (@markgurman)
11:53 PM • Jun 1, 2025
🤝 Samsung's alliance with Perplexity underscores a strategic pivot in the mobile industry, as manufacturers seek to enhance user experiences through diversified and integrated AI technologies.
🆕 Updates
Introducing Conversational AI 2.0
Build voice agents with:
• New state-of-the-art turn-taking model
• Language switching
• Multicharacter mode
• Multimodality
• Batch calls
• Built-in RAGNow fully enterprise-ready with HIPAA compliance, EU data residency, and robust
— ElevenLabs (@elevenlabsio)
7:03 PM • May 30, 2025
Significant progress in AI and Robotics this week.
So, I summarized everything from Google DeepMind, Anthropic, Figure, DeepSeek, Sakana AI, Perplexity, ETH Zurich, Unitree, Physical Intelligence, and more.
Here's everything you need to know and how to make sense out of it:
— Brett Adcock (@adcock_brett)
4:03 PM • Jun 1, 2025
Manus slides can now be exported as Google Slides!
Create with Manus, collaborate everywhere!
— ManusAI (@ManusAI_HQ)
1:44 PM • Jun 2, 2025
📽️ Daily Demo
Financial Researcher/Analyst is now a prompt on Perplexity Labs
— Aravind Srinivas (@AravSrinivas)
10:42 PM • May 29, 2025
One weird trick to stop and correct hallucinations
— Garry Tan (@garrytan)
9:34 PM • May 31, 2025
Flux Kontext via @replicate
Kontext just put every photo restoration company out of business. This took 6 seconds.
— Adam Hails (@Clearstory3D)
6:51 PM • May 30, 2025
It’s over.
FLUX 1 Kontext edits AI images live
Keeps your character and text consistent, pixel-perfect tweaks in seconds 🤯
10 wild examples + how to try: 👇
1. Change hair style
— Min Choi (@minchoi)
3:43 PM • May 31, 2025
🗣️ Discourse
Veo 3 is out of control right now!
13 wild examples, prompts included 🧵👇
1. A chimpanzee operating a jackhammer on a city sidewalk
— TechHalla (@techhalla)
10:48 PM • May 31, 2025
The demand for Veo 3 has been off the charts! Millions of videos have been generated in the past few days alone, now available on mobile, and in more countries (including the UK! 🇬🇧). Try it out in the @GeminiApp (and the Flow tool for filmmakers at )
— Demis Hassabis (@demishassabis)
11:46 PM • May 30, 2025