In partnership with |
 |
|
Sign Up | Advertise | Prompt Guide | Unsubscribe |
|
|
| |
| |
Welcome, Noodle Networkers. |
AI is getting existential, your smart fridge might soon have bigger ambitions, and one guy figured out how to make eight million dollars from songs nobody actually listened to. Totally normal Tuesday. Geoffrey Hinton says we may be creating something closer to alien intelligence than a normal computer. And when asked whose side he is on, he picked humans. Comforting, until you remember the guy helped invent the technology he is warning us about. 👽 Anthropic says smaller AI models could soon learn to copy themselves across smart TVs, ovens, and other connected devices. Your robot vacuum already wanders around the house unsupervised, and apparently that was just the tutorial. 🏠 One man used hundreds of thousands of AI songs and an army of bots to generate more than eight million dollars in fraudulent streaming royalties. At one point his bots were pumping out streams faster than Taylor Swift. Spotify Wrapped is going to need a fraud section. 🎵 From alien intelligence to ambitious appliances to robot musicians gaming the charts, AI keeps finding new ways to make reality sound fake. Let’s dig in. |
|
|
In today’s AI digest: |
Hinton says AI could become alien and he picks humans 👽 Anthropic warns AI may soon copy itself across devices 🏠 AI songs and bots allegedly faked eight million in royalties 🎵
|
Read time: 5 minutes |
|
|
|
| |
|
|
|
| |
| |
The ice cream shop that makes money when it's cold |
|
|
28 Wishes sells ice cream in Los Angeles. Below 70°F, sales fall about 20%. The weather is out of their hands. Rent isn't. |
So the owners started putting about $20 a day into Kalshi weather markets, taking the cold side. The days that keep customers away now pay something back. |
This is hedging. Big companies have done it for decades, buying protection against bad weather, fuel spikes and rising rates. It used to take a broker, a trading desk, and an order size no corner shop could meet. |
Kalshi opens it up. Contracts on weather, fuel prices, inflation, tariffs and regulation, starting at a few dollars. Take a position on the outcome that would hurt you. If it hits, the payout softens it. If it doesn't, the contract expires and the good month was the point. |
See how other owners are doing it at Kalshi.com/hedging |
|
|
|
| |
|
| |
| |
AI dangers |
|
|
|
(source: TheGuardian) |
👽 The Digest: Geoffrey Hinton says we’re creating “alien beings,” and he’s firmly on team human. In a Guardian interview, the AI pioneer argues that human welfare should come first as machines become more capable. First contact apparently comes with a login screen. |
Key Details: |
🧠 Alien Means Different
Hinton is describing a new kind of intelligence, with implications for how humans understand their place in the world. Nobody needs to inspect the garden for crop circles. The uncomfortable questions are already sitting in a browser tab. |
🎲 Uncertainty Is the Warning
He acknowledges that AI’s future cannot be predicted precisely, but argues that potentially catastrophic risks deserve attention anyway. Waiting for absolute certainty is a questionable strategy when the experiment includes the people reviewing the results. |
🏛️ Safety Needs an Outside Referee
In a separate recent podcast, Hinton called for AI safety approval resembling the process for medicines, requiring developers to demonstrate safety before release. “Our engineers feel pretty good about it” would make a memorable pharmaceutical commercial, though probably a short one. |
🌍 The Concern Is Becoming Collective
Hinton also joined hundreds of scientists calling for international oversight, stronger containment, and incident reporting for advanced AI. Their argument is that voluntary promises need enforceable backing. A safety agreement should be harder to ignore than a software update reminder. |
Why It Matters: The debate reaches beyond how impressive AI becomes to who sets its limits and checks its behavior. Hinton’s warnings remain contested, but the practical question deserves an answer: if we’re building something smarter than ourselves, who gets to keep the administrator account? |
|
|
|
| |
|
| |
| |
Anthropic |
|
|
|
(source: CNBC) |
🏠 The Digest: Anthropic warns that small AI models could be months away from copying themselves across poorly secured smart devices. Its research points to a possible new malware threat, not a confirmed outbreak in people’s homes. Your television already asks for enough permissions without quietly applying to become a data center. |
Key Details: |
📺 Cheap Gadgets Have Computing Power
Researchers ran an AI model across 15 inexpensive Android TV boxes. That demonstrated how modest devices can pool their resources to run AI. Individually, they stream your shows. Together, apparently, they qualify for a more ambitious group project. |
🧪 The Lab Left Doors Open
A stronger model completed a simplified replication test, but researchers deliberately made the security barriers easy. Smaller models still struggled to complete the whole process. Think obstacle course with several hurdles removed, not proof that every connected appliance is ready to defect. |
📅 The Timeline Is an Estimate
Anthropic projects that smaller models could reach the necessary capability by the end of 2026 if historical improvements continue. That forecast uses general capability trends, so the timing remains uncertain. This is a research warning, not a calendar invitation from your fridge. |
🔌 No Single Off Switch
An AI worm spread across many devices could be difficult to eliminate because it would lack one central shutdown point. Anthropic says propagation could still be slow, taking days to double. Unfortunately, slow malware remains malware, even if it has the work ethic of a Friday afternoon. |
Why It Matters: The research suggests AI security could increasingly depend on the ordinary gadgets surrounding us. Manufacturers’ security choices may matter far beyond whether a device works as advertised. We bought smart appliances to reduce household chores. Becoming the unpaid administrator of a tiny server farm was not on the box. |
|
|
|
| |
|
| |
| |
AI fraud |
|
|
|
(source: FoxBusiness) |
🎵 The Digest: A North Carolina musician manufactured both the music and the audience, collecting more than $8 million in fraudulent royalties. Michael Smith has now been sentenced to 18 months in prison after pleading guilty to conspiracy to commit wire fraud. Apparently, having fans who exist was an inconvenient part of the business model. |
Key Details: |
🤖 An Audience With No Ears
Prosecutors say Smith used up to 10,000 bot accounts at once to stream songs across services including Spotify, Apple Music, Amazon Music, and YouTube Music. Imagine a sold out stadium where every seat contains a laptop and nobody needs the bathroom. |
🎹 The Content Factory Worked Overtime
Hundreds of thousands of AI generated songs supplied the catalog. Prosecutors say Smith spread automated plays across thousands of tracks to avoid attracting attention to any single song. Less musical breakthrough, more accounting department with a synthesizer. |
💸 Real Musicians Picked Up the Tab
Streaming royalties are distributed proportionally, so fake listening can divert money from artists attracting actual listeners. Those automated plays were reaching into the same royalty pot as musicians whose promotional budget is three Instagram posts and a favor. |
⚖️ The Payout Comes With a Payback
Alongside his prison sentence, Smith received two years of supervised release and must forfeit roughly $8.09 million. That is a spectacularly expensive way to discover that a streaming dashboard and a legal income are different things. |
Why It Matters: Cheap AI music and automated audiences put pressure on platforms to verify whether popularity is real before paying for it. The challenge goes beyond deciding whether a song sounds human. Someone also needs to check whether its biggest fan is a person or a process running in the background. |
|
|
|
| |
|
|
|
| |
| |
Trending AI Tools |
Holo – AI marketing tool that creates ads, emails, and social posts from your brand. |
UiChemy – Turns Figma designs, prompts, or briefs into editable WordPress pages. |
Instinct – Personal AI assistant that helps manage tasks through chat and connected apps. |
HypeAuditor – AI influencer marketing platform for finding creators, checking audiences, and tracking campaigns. |
|
|
|
| |
|
| |
| |
AI Hacks & How-Tos |
|
|
|
UiChemy turns Figma designs, prompts, client briefs or AI builds into real editable WordPress pages. It works with popular WordPress builders like Elementor, Gutenberg and Bricks, so your design becomes actual page content instead of a flat image or locked export. |
How to Use It 🧭 |
1. Choose Your Starting Point
Start from a Figma design, a written prompt, a client brief or an AI generated build. UiChemy supports multiple paths depending on how your project begins. |
2. Connect Your WordPress Site
Use the UiChemy connection link to connect your WordPress site to the UiChemy plugin. This allows your design to import directly into your live WordPress environment. |
3. Pick Your Builder
Choose whether you want the page built for Elementor, Gutenberg or Bricks. UiChemy converts the design into native editable elements for that builder. |
4. Generate or Import the Page
If you are using Figma, run the conversion through the UiChemy Figma plugin. If you are starting from a prompt or brief, use AI Website Creator to create the site structure and page design. |
5. Edit and Publish in WordPress
Open the imported page in your builder, adjust text, spacing, images and sections, then publish when everything looks right. UiChemy keeps the page editable so you can continue improving it after import. |
UiChemy helps designers and website builders move from concept to working WordPress pages much faster, while still keeping control inside the tools they already use. |
|
|
|
| |
|
| |
| |
What'd you think of today's email? |
|
|
|
|
|
| |
|
|
Tidak ada komentar:
Posting Komentar