- The Current ⚡️
- Posts
- ChatGPT Now Integrates With Cloud Storage & More
ChatGPT Now Integrates With Cloud Storage & More
Also, Runway making a play for Hollywood partnership

⚡️ Headlines
🤖 AI
Claude Explains: Anthropic’s Conversational AI Demystified – Anthropic launches an interactive explainer to clarify how its Claude AI model functions and interprets user prompts. [Anthropic]
Mistral Releases “Code” — A Vibe-Centric AI Coding Client – Mistral unveils a new AI-assisted coding environment designed to support real-time, collaborative programming. [TechCrunch]
Google Pauses “Ask Photos” AI Rollout Amid Privacy Concerns – Google temporarily halts the rollout of its AI-powered image search tool for Google Photos due to safety evaluations. [The Verge]
AI Hacker Agents Are Coming — and You’re Not Ready – Security experts warn of AI agents autonomously launching sophisticated cyberattacks without human oversight. [Wired]
AI Summarizers Tested on Books and Docs Still Miss the Mark – A review of top AI summarizers reveals significant flaws in accurately distilling complex texts like books and lengthy documents. [Washington Post]
Figma Debuts Dev Mode MCP Server Beta to Enhance AI Workflows – Figma introduces a beta version of its Dev Mode MCP Server, streamlining AI tool integration for collaborative design and development. [The Verge]
Gemini 2.5 Pro Preview Shows Enhanced Vision and Reasoning – Google previews Gemini 2.5 Pro, touting major improvements in multimodal understanding and contextual reasoning. [9to5Google]
Hollywood Grapples with Generative AI’s Impact on Creativity – Writers and studios debate generative AI’s role in storytelling as the technology infiltrates film and TV production. [Vulture]
OpenAI Details Chinese Influence Operation Disrupted via ChatGPT – OpenAI reveals it disrupted a coordinated effort to manipulate public discourse using ChatGPT in English and Chinese. [NPR]
🦾 Emerging Tech
Meta Pursues Hollywood Partnerships for New VR Headset – Meta aims to team up with studios to create immersive content tailored for its upcoming virtual reality hardware. [WSJ]
Amazon Readies Humanoid Robots for Package Delivery Trials – Amazon is preparing field tests of humanoid robots designed to assist with home package delivery. [The Information]
Circle Soars on NYSE Debut After Above-Range IPO Pricing – Stablecoin firm Circle sees a strong market entry with its IPO pricing exceeding expectations. [CNBC]
Volvo Adds AI-Powered Seatbelt Guidance in EX60 SUV – Volvo’s EX60 introduces AI-assisted features that advise users on proper seatbelt placement and comfort. [Automotive News Europe]
🤳 Social Media
X Adds Community Notes Labels to More Content, Per Musk – Elon Musk announces expanded use of Community Notes on X to combat misinformation and provide crowd-sourced context. [Axios]
Xiaohongshu’s $26B Valuation Boosts Chinese Tech Investment – The valuation surge of Chinese social app Xiaohongshu provides momentum to GSR Ventures’ investment portfolio. [Bloomberg]
🔬 Research
Researchers Develop New Technique to Create Synthetic Embryos – Scientists succeed in growing model human embryos from stem cells without using eggs or sperm. [BBC]
⚖ Legal
Reddit Sues Anthropic Over AI Training Data Usage – Reddit alleges that Anthropic unlawfully used Reddit data to train its AI models without permission. [WSJ]
OpenAI Says Forced Logging of ChatGPT Use Threatens Privacy – OpenAI argues that a court order mandating retention of all ChatGPT logs would create severe privacy risks. [Ars Technica]
Apple Loses Appeal to Delay App Store Overhaul in Epic Case – A court denies Apple’s request to postpone mandated changes to its App Store following the Epic Games antitrust ruling. [Reuters]
States Challenge Federal Plan to Centralize AI Regulation – State legislators oppose a federal proposal that would limit local authority over AI governance. [StateScoop]
Apple Shared Push Notification Metadata with Governments – Apple confirmed it has handed over metadata from push notifications to various governments in response to legal requests. [404 Media]
🎱 Random
IRS Direct File Open Source Faces Industry Pushback – Intuit opposes the IRS’s open-source tax filing platform, fearing it could disrupt the commercial tax prep industry. [404 Media]
23andMe Sets New Auction as Ex-CEO Submits $305M Bid – Genetic testing firm 23andMe receives a revised buyout offer from its former CEO, prompting a fresh round of bidding. [Bloomberg]
🔌 Plug-Into-This
OpenAI has unveiled new enterprise features for ChatGPT, including integrations with cloud storage platforms and a "record mode" for meetings. These enhancements aim to streamline workflows by allowing users to access, search, and summarize data across services like Google Drive, Dropbox, Box, SharePoint, and OneDrive directly through ChatGPT.

Cloud Storage Integration: ChatGPT now connects with major cloud services, enabling users to retrieve and interact with documents and spreadsheets stored across various platforms.
Meeting Recording and Transcription: The new "record mode" allows for automatic recording and transcription of meetings, generating time-stamped notes and suggested action items.
Deep Research Capabilities: Users can leverage connectors for tools like HubSpot and Linear to compile detailed research reports, integrating data from multiple sources.
Model Context Protocol (MCP) Support: Adoption of MCP facilitates standardized connections between ChatGPT and external tools, enhancing its utility in professional settings.
Enterprise Adoption Growth: OpenAI reports an increase in its enterprise user base, indicating growing reliance on ChatGPT for business operations.
Woo hoo! I'm ludicrously excited about the @OpenAI announcement today that ChatGPT now has a connector built right in for HubSpot to bring data in for deep research.
Super cool to see the HubSpot logo inside the world's most popular AI app.
— dharmesh (@dharmesh)
8:45 PM • Jun 4, 2025
🗃️ These updates position ChatGPT as an even more integral tool in enterprise environments, reflecting OpenAI's strategy to embed AI deeper into daily business operations as an all-in-one productivity solution.
Cris Valenzuela, CEO of Runway, discusses the company's advancements in AI-driven video generation and its implications for the film industry. He emphasizes collaboration with major studios and addresses concerns about copyright and the role of AI in creative processes.

AI Video Generation: Runway enables users to create videos using reference images and text prompts, evolving from basic outputs to high-quality 4K videos.
Industry Collaboration: The company is working with major studios like Lionsgate and AMC Networks to integrate AI tools into traditional filmmaking workflows.
Copyright Considerations: Valenzuela acknowledges legal challenges regarding the use of copyrighted materials in training AI models and advocates for fair compensation and transparency.
Democratizing Content Creation: Runway aims to make high-quality video production accessible to a broader audience, reducing barriers to entry in the creative industry.
Balancing Innovation and Tradition: The company positions itself as an enabler rather than a disruptor, seeking to enhance rather than replace traditional filmmaking techniques.
That Vulture piece on AI in Hollywood is bleak.
vulture.com/article/genera…
— Darren Mooney (@Darren_Mooney)
8:40 PM • Jun 4, 2025
🎬 Runway is coming for Hollywood, bottomline. Regardless of whether they choose to be nice about it, the film industry was already teetering on the point of collapse and companies like Runway are poised to eat up the pieces.
Demis Hassabis, CEO of Google DeepMind, shares his vision of artificial general intelligence (AGI) as a transformative force that could lead to a more cooperative and less self-centered human society. He discusses the potential of AGI to address global challenges and the importance of ethical considerations in its development.

AGI Timeline: Hassabis estimates a 50% chance of achieving AGI within the next 5 to 10 years, envisioning systems that replicate all human cognitive abilities.
Global Problem-Solving: He believes AGI could help solve critical issues like disease, climate change, and energy scarcity, leading to "radical abundance."
Behavioral Impact: The shift from zero-sum to non-zero-sum scenarios facilitated by AGI could encourage more altruistic human behavior.
Ethical and Regulatory Emphasis: Hassabis stresses the need for international cooperation and regulation to mitigate risks associated with AGI, including misuse and safety failures.
Economic Considerations: He calls for the development of new economic theories to address the changes brought about by a post-AGI world.
My (brief) thoughts on the definition of AGI, why we aren't there yet, and what is missing. Will be writing a short and succinct essay on this important topic soon.
— Demis Hassabis (@demishassabis)
8:09 AM • May 22, 2025
🤔 Hassabis basically believes that if AI can solve big problems and provide plenty for everyone, people might become more generous and less focused on competing for limited resources. Feels pretty hopeful, and I guess we’ll find out one way or another, as this industry shows no signs of stopping.
🆕 Updates
Introducing Modify Video. Reimagine any video. Shoot it in post with director-grade control over style, character, and setting. Restyle expressive performances, swap entire worlds, or redesign the frame to your vision. Shoot once. Shape infinitely.
— Luma AI (@LumaLabsAI)
2:34 PM • Jun 4, 2025
Take the reins — Reorder your track. Remix the vibe. Split it into stems. Upload full songs.
HERE’S WHAT’S NEW:
🎼 Upgraded Song Editor: Make any edit you can imagine. Reorder, rewrite, and remake your track section by section—right from the waveform.
🧬 Stem Extraction:
— Suno (@SunoMusic)
9:05 PM • Jun 3, 2025
Introducing Eleven v3 (alpha) - the most expressive Text to Speech model ever.
Supporting 70+ languages, multi-speaker dialogue, and audio tags such as [excited], [sighs], [laughing], and [whispers].
Now in public alpha and 80% off in June.
— ElevenLabs (@elevenlabsio)
6:14 PM • Jun 5, 2025
📽️ Daily Demo
Aria Gen 2 glasses mark a significant leap in wearable technology, offering enhanced features and capabilities that cater to a broader range of applications and researcher needs. We believe researchers from industry and academia can accelerate their work in machine perception,
— AI at Meta (@AIatMeta)
4:00 PM • Jun 4, 2025
🗣️ Discourse
Good post from @balajis on the "verification gap".
You could see it as there being two modes in creation. Borrowing GAN terminology:
1) generation and
2) discrimination.
e.g. painting - you make a brush stroke (1) and then you look for a while to see if you improved the— Andrej Karpathy (@karpathy)
4:46 PM • Jun 4, 2025
This will probably get settled; Reddit just suing to increase the pressure?
The AI companies, including Anthropic, don't want to set a new precedent that they must respect Terms of Service... Prior claims (e.g. Genius v Google) were deemed to be preempted by © law and dismissed.
— Alex J. Champandard 🌱 (@alexjc)
6:21 PM • Jun 4, 2025
With less than five days of notice, Anthropic decided to cut off nearly all of our first-party capacity to all Claude 3.x models. Given the short notice, we may see some short-term Claude 3.x model availability issues as we have very quickly ramped up capacity on other inference
— Varun Mohan (@_mohansolo)
10:52 PM • Jun 3, 2025
I'm speechless...
This is a major step-change in AI generated voice (TTS)
Elevenlabs just released v3 as a public alpha and you can try it now for 80% off during all of June 💸
Let's unpack this 👇
— Linus Ekenstam (@LinusEkenstam)
6:20 PM • Jun 5, 2025
Cannot believe to what degree LLM workflows are totally held back by insufficient infrastructure and overloaded servers
Token limits too low, random latency ruining voice calls, unannounced downtime
Good problems to have but also wow, we are still so early
— Garry Tan (@garrytan)
6:19 AM • Jun 6, 2025