ChatGPT connectors for Google Drive, Dropbox, SharePoint, and Box are now available to Pro users (excluding EEA, CH, UK) in ChatGPT outside of deep research.
Perfect for bringing in your unique context for everyday work.
CEXes are finally taking note and going permissionless: @coinbase, @binance, and @Bybit_Official just kicked off a “CEX-on-chain” market grab:
• Coinbase: DEX trading + Verified pools
• Binance: Binance Alpha (curated + sandboxed tokens)
• Bybit: Byreal RFQ+CLMM
- - - - -
1. Coinbase | DEX Trading + Verified Pools
• @base-native routing for any on-chain tokens, wrapped with KYC-verified pools that institutions can actually touch.
• Breadth of DeFi + TradFi regulation standards, all inside the app millions already use
2. Binance | Binance Alpha
• A curated sandbox of early-stage on-chain tokens (across Ethereum, BSC, Solana, Sui, Sonic)
• Discovery platform where users can purchase them without leaving the Binance app
• Pre-listing spotlight, as some tokens will graduate to a proper spot listing
3. Bybit | Byreal
• @solana-powered RFQ + CLMM router built to replicate CEX spreads on-chain
• Fair-launch vaults and yield layers baked in to keep capital sticky from day one
- - - - -
Closing
These show how CEXes are adapting:
• Building solutions to bridge CeFi and on-chain DeFi
• Adding verification/trust layers to DeFi (common concerns for most, especially hacks and rug pulls) --> get people on-chain within the comfort of a web2 app
• Hybrid solutions that leverage their brand reputation while benefiting their current user base
• Cannot ignore the volumes and traction that @HyperliquidX is having
• Adheres to the Internet Capital Market thesis: Decentralized, permissionless marketplaces where capital flows directly to builders, creators, and ideas --> this will be instant and global
• Token price discovery are taking place on-chain (just look at your memes, AI tokens)
Builders need on-chain rails to stay discoverable; traders need them to stay liquid.
Rather than fight DeFi's growth, they're positioning themselves as (potential) gateways to the ever growing on-chain ecosystem.
Projects Update 📝
We’re adding more capabilities to projects in ChatGPT to help you do more focused work.
✅ Deep research support
✅ Voice mode support
✅ Improved memory to reference past chats in projects
✅ Upload files and access the model selector on mobile
BREAKING 🚨: Project Mariner is being rolled out to Gemini Ultra subscribers!
It works more like an Operator but has access to your open tabs in Chrome.
🔥Veo 3 keeps growing like crazy. To keep up, we’re introducing Veo 3 Fast in @GeminiApp and Flow. It’s >2x faster, has the same 720p resolution, and a bunch of serving optimizations. The big headline: we can serve more of it, even for the Yetis!
How to get started:
1) Get a Google AI plan: https://t.co/KgFOQbr08u
2) Use @GeminiApp for quick + everyday access: https://t.co/03ZZ0aoRbI
3) Use Flow for AI filmmaking: https://t.co/zd7SeFVnY2
More details in 🧵
ChatGPT can now connect to more internal sources & pull in real-time context—keeping existing user-level permissions.
Connectors available in deep research for Plus & Pro users (excl. EEA, CH, UK) and Team, Enterprise & Edu users:
Outlook
Teams
Google Drive
Gmail
Linear
& more
Introducing Manus video generation.
Manus transforms your prompts into complete stories—structured, sequenced, and ready to watch. With a single prompt, Manus plans each scene, crafts the visuals, and animates your vision. From storyboard creation to concept visualization—your ideas become living videos in minutes.
Early access for Basic, Plus, Pro Members. Coming soon for everyone!
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.
@OpenAI is testing a new feature that would let users sign into third-party apps using their ChatGPT accounts.
Like how you use Google to sign in across apps.
Devs can preview how it works in Codex CLI and apply to take part in the rollout via a short form.
The idea is to make ChatGPT accounts portable across tools, opening up easier user access and deeper integrations.
It’s still early, but it hints at a broader identity layer emerging around OpenAI’s platform.
@AnthropicAI just released a new voice mode for Claude on iOS & Android.
It adds spoken interaction and deeper control of Google Workspace apps.
Users can now ask Claude to check Gmail, summarise Google Docs, or read calendar entries aloud with multiple voice options.
https://t.co/AVNABveZG0
Google added a way for developers to see inside how Gemini thinks.
If you’re using “thinking budgets” in the Gemini API, you can now turn on thought summaries.
These are short, human-readable explanations of the model’s reasoning.
They’re free, optional, and only need one line of code to switch on.
The summaries don’t change the model’s output and won’t affect your billing.
& it’s still experimental, but custom styles are coming.
This is an early step toward making AI reasoning more transparent.
& it can help devs debug, interpret, and trust what the model is doing under the hood.
https://t.co/TOKMNYXa0n
We just rolled out “thought summaries” in the Gemini API, now you can see what the model is thinking and make use of that info!
A thread with the details and request for feedback 🧵
That's a really interesting thought! Demis Hassabis says that hallucinations aren't necessarily bad. They helped AlphaEvolve develop the new algorithm. That's surprising!
Should we rethink hallucinations? As creativity?
@OpenAI is building a tiny AI device that could replace your screen.
They just locked in a $6.5B deal to acquire Jony Ive’s hardware startup, io.
They want to build a new kind of personal computing device, something small, screenless, and always on.
@sama calls it the “third core device” after the iPhone & MacBook Pro.
It’s meant to live in your pocket or on your desk, quietly listening, interpreting your surroundings, and helping you interact with AI in real time.
It won’t be a phone. One of the goals is to help us rely less on screens.
The first version is set to launch in late 2026.
It’ll be slightly bigger than the now-discontinued Humane AI Pin.
Maybe something similar to the iPod Shuffle, for those who remember it.
Design is being led by Ive’s firm, LoveFrom, with support from a team of ex-Apple engineers now working at OpenAI.
The vision includes a full family of integrated, Apple-style AI devices & OpenAI wants to ship 100 million+ of them.
The wait is over... for your responses that is. Now, when you ask a question in the Chat box, the reply will appear as it's generated (instead of all at once). This will reduce wait time *and* make your conversation flow smoother.
Hey! There's no such thing as small launches.
Introducing the next generation: Claude Opus 4 and Claude Sonnet 4.
Claude Opus 4 is our most powerful model yet, and the world’s best coding model.
Claude Sonnet 4 is a significant upgrade from its predecessor, delivering superior coding and reasoning.
The last time humans built a new civilisation was during the Enlightenment.
It gave us the world we live in today: democracy, capitalism, industrial tech.
It started with better communication tools: printed books, public mail... which allowed thinkers to exchange ideas.
Now it’s happening again.
Today, we have the internet. We also have a new form of intellectual network: @Substack.
We’re entering a new era shaped by tech, AI, clean energy, and biology.
And with that, we need to rethink our systems from the ground up, from capitalism to governance.
That work won’t start in the government.
History doesn't repeat, but it does rhyme.
The Enlightenment changed the world with letters.
This one might do it with posts.