In partnership with
Welcome Automaters, 👋
Imagine opening your inbox to find a massive financial institution practically begging its customers to stop doing something wildly unsafe.
Well, Bank of America just had to send out an official, emergency-vibes email titled "Protecting your finances in the AI World." The core message? Kindly stop handing over your online banking usernames and passwords to your shiny new AI agents.
You can practically taste the executive frustration behind that send button!
Here’s the tea: as autonomous AI tools get smarter, people are eagerly handing over the keys to their bank accounts so these digital helpers can handle their finances. But standard bank security policies are screaming red alert.
Why This Is a Huge Disaster Waiting to Happen
Handing your secret login info to a bot exposes you to three massive risks:
Unsupervised Shopping Sprees: Your AI could execute a transaction without your explicit approval, leaving you with an empty checking account.
Instant Account Lockouts: Bank fraud departments don't care that it was just your favorite digital assistant logging in from a strange server; they will think your account is being raided and shut it down immediately.
Massive Data Leaks: A single security hole in that third-party AI platform could broadcast your financial credentials straight to cybercriminals.
This drama highlights a huge friction point in fintech today. While major banks and payment processors are actively trying to figure out how to safely enable agentic payments (where your AI automatically buys products at the best price or optimal moment), current systems simply aren't ready for raw password sharing.
Here’s the real kicker: If your money vanishes because your AI assistant got hacked or went rogue, you are on the hook.
Most bank account agreements explicitly state that if you voluntarily give out your credentials to anyone or anything, you shoulder the blame for the consequences. Bank of America put it bluntly in their notice: "If you do share your credentials, keep in mind that you’re responsible for activity resulting from that access."
The Takeaway: for the tenth time, keep your logins to yourself, let the banks figure out secure API access, and don't let a bot ruin your credit score!
Here's what we have for you today
🤯 How Coder Emir Karşıyakalı Brought Modern AI to the Nokia 6300
So we have some delicious retro tech gossip that’s going to make your inner 2000s teenager absolutely scream!
Cast your mind back nearly two decades ago. Steve Jobs was up on stage changing the world with the very first iPhone. But while Apple was taking over, Nokia dropped a sleek, unassuming little metal brick called the Nokia 6300. It quickly became an absolute juggernaut seller because it was cheap, practically indestructible, and drop-dead gorgeous.
Fast forward to 2026, and that iconic little dumbphone is essentially a relic on borrowed time.
For starters, it relies on ancient 2G cellular networks that carriers are aggressively shutting down. Even worse? Its built-in browser stopped supporting modern internet security standards years ago, leaving it completely locked out of almost 80% of today's web.
Most people would throw it in an electronic waste bin. But one brilliant developer looked at this piece of tech history and said: "Hold my coffee, I'm putting generative AI on this thing."
So yeah, meet Turkish entrepreneur Emir Karşıyakalı.
Back in late September, Emir took to X to unveil a wild side project he vibe-coded: an app called Claude S40. His initial build allowed a original Nokia 6300 to talk directly to Anthropic’s Claude chatbot using a lightweight Java and Go setup.
Since then, he expanded the project to work with all the heavy hitters—including ChatGPT, Gemini, and Grok. He officially renamed the platform Aikon (which is literally Nokia spelled backwards) to celebrate its massive upgrade.
How Does This Actually Work?
You can’t just download ChatGPT from an app store on a device with 7.8 megabytes of storage! Here’s the lowdown on how Emir pulled it off:
You run a custom Go server on a modern computer that holds your personal API keys for OpenAI, Google, or Anthropic.
The server talks to the AI, takes those massive responses, and chops them up into tiny 8kB data packages that the phone’s tiny brain can actually process.
Because you are tapping directly into developer APIs, you do need to set up standard API token billing with your chosen platform.
And get this: It’s surprisingly loaded with features
You might assume this is just a quick gimmick, but Emir built this thing like a total powerhouse. It’s got:
Dedicated Reading Mode: you can hit a keypad shortcut to format text so it actually looks readable on a tiny 240x320 screen.
Offline Saving: you can save full AI responses straight to your microSD card for later browsing.
Multimodal Capabilities: you can also send photos directly to your chatbot for image analysis using the phone's camera.
Voice Notes: Record up to 30 seconds of audio on your Nokia, and OpenAI's Whisper model will translate your speech into text.
Global Support: The server stores conversation threads for up to 30 days and supports 8 major languages (including English, French, Spanish, German, and Turkish).
Best of all? Emir made the whole thing completely open-source, meaning anyone can grab the code and tweak it!
The Bad News for American Readers
Before you run off to eBay to buy a mint-condition silver 6300, there’s a massive catch.
If you live in the United States, you cannot use this app. The country’s final surviving 2G GSM network was officially switched off by T-Mobile this past August, leaving these older devices completely disconnected from cellular networks.
However, if you are living outside North America where 2G is still alive and kicking, you can head over to GitHub right now, follow the setup instructions, and flex modern AI on a 20-year-old gadget!
PS: You can read the full report here.
What the People Building AI Actually Read
The people building AI aren't only reading about AI. Sam Altman's pick for getting more done is Zero to One, and it has nothing to do with machine learning.
Shortform breaks down books like The Beginning of Infinity and Thinking, Fast and Slow into the key ideas, the context behind them, and what to do with them.
Get the ideas behind the breakthroughs in about 15 minutes, then decide which ones deserve your whole weekend.
🧱 Around The AI Block
🫣 Fashion's AI era: face swaps and digital clones.
😨 DeepSeek AI model shows safety gaps in independent tests.
🤷 Microsoft offers employees money to exit company.
👍 Opera introduces a new web standard to let users deploy AI locally in their browser.
😕 Frontier AI labs say their models refuted the Exponential Time Hypothesis. What does that mean?.
🤦♂️ ‘Meaningless as a laptop screensaver’: reviewers slate Musk’s AI-made Odyssey film.
🤩 Google’s Genie 4 world model adds audio, more realistic worlds, and real-time steering.
💼 Meta AI director Rohit Patel leaves to take on Microsoft Office and Google Workspace.
🤔 Who controls powerful AI? Anthropic has a plan for long after its leaders depart.
🛠️ Trending Tools
For the Brand Builders: Shutterstock AI Generator is built on tech from OpenAI and Gemini, to allow you generate high-end visuals that are actually commercially licensed.
For the Browser Power-Users: Gyde is a Chrome extension that turns any web workflow into a step-by-step video and screenshot tutorial automatically. Just do the task once, and the AI handles the documentation.
For the High-Stakes Communicators: Mailtrap provides a "safety net" with high-inbox delivery rates and separate streams for bulk vs. personal AI-generated sends, ensuring your automated outreach actually gets seen.
🤖 AI Workout Of The Day: How to Lock Your ChatGPT Chats Before Someone Sees
Let’s be honest: handing your phone to a friend or scrolling through ChatGPT in public can feel like a high-stakes extreme sport. One wrong scroll, and boom, everyone knows what you asked AI at 3:00 AM.
Well, OpenAI finally gets the hint. They’re rolling out a brand-new feature called Locked Chats, and it’s here to save us all from public embarrassment.
How the New "Locked Chats" Vault Works
Once you switch this feature on, you get total control over where your private exchanges show up.
When enabled, your chosen chats will completely disappear from:
Your regular chat history
Your pinned chats
Your global conversation search
Instead, all your hidden conversations get swept away and neatly tucked under a brand-new project folder titled "Locked Chats."
To open the vault, you simply enter a 4-digit PIN. If you are using an iPhone or Android device, you can bypass the PIN entirely using Face ID or your phone’s standard biometric check.
Now, before you go locking up your entire history, OpenAI dropped a few serious disclaimers on their support page that you cannot ignore:
It is NOT End-to-End Encryption: This feature is built to keep prying eyes off your screen, period. It will not protect you if you get into legal trouble or breach your company's workplace policy.
It Does NOT Stop Model Training: Locking a chat does not automatically opt you out of data retention or training settings. You still need to manage those in your privacy menu.
The Self-Destruct Warning: Pay attention here. If you forget your 4-digit PIN and reset it, OpenAI will permanently delete every single locked chat and uploaded file. Support cannot recover them for you. Consider yourself warned!
Who Gets It & How to Use It
OpenAI is rolling this out gradually, so if your app does not have it yet, give it a few days!
Supported Plans: Available for Free, Go, Plus, and Pro users.
Voice Mode Compatible: You can use Voice Mode inside locked chats on both iOS and Android.
The Catch: You cannot lock your Codex coding conversations behind this feature just yet.
Go grab the update, lock down your wild late-night prompts, and scroll in peace!
💡 Prompts to Try:
-------------------Is this your AI Workout of the Week (WoW)? Cast your vote!
That's all we've got for you today.
Did you like today's content? We'd love to hear from you! Please share your thoughts on our content below👇
What'd you think of today's email?
Your feedback means a lot to us and helps improve the quality of our newsletter.
