SMRTR AI

AI stories

Get hand-picked daily summaries of the best, most informative AI articles from around the web. Browse recent summaries, then jump to the original source when you want the full article.

Latest story• Feb 1, 2026

Deepfake job seeker applied to work for an AI security firm

Deepfake technology collided with old-fashioned hiring practices when cybersecurity expert Jason Rebholz found himself interviewing what appeared to be a completely fabricated job candidate. The CEO of AI security firm Expel watched red flags pile up during a virtual interview: an anime profile picture, an AI-generated resume, and urgent follow-up messages that screamed scammer playbook. But when the candidate...

Read summary Original source

Latest stories

Two kinds of AI users are emerging. The gap between them is astonishing

Two types of AI users are emerging: power users with advanced tools like Claude Code achieve massive productivity gains, while enterprise workers stuck with inferior tools like Microsoft Copilot lag behind significantly.

Read summary Original

How does AI impact skill formation?

A recent Anthropic study found that people learning Python's Trio library using AI performed worse on knowledge retention tests. However, half the AI users inexplicably spent time manually retyping generated code rather than...

Read summary Original

Major AI Companies Aren’t Even Pretending to Make Money

Major AI companies are raising billions despite having no revenue or clear business plans. These firms operate on hype rather than financial reality, with some rejecting multi-billion dollar offers.

Read summary Original

Too much open-source AI is exposing itself to the web

Security researchers found over 175,000 internet-exposed Ollama AI systems across 130 countries, creating massive vulnerabilities. These systems lack proper security controls and use identical configurations, making them...

Read summary Original

Why is OpenAI so stingy with ChatGPT web search?

OpenAI deliberately hides ChatGPT's web search feature despite the AI's accuracy issues without it. This contradicts their heavy user acquisition spending while restricting tools that improve performance.

Read summary Original

SMRTR AI

Get the next batch of curated stories in your inbox.

This archive is built from SMRTR newsletter stories. Subscribe for hand-picked stories without the extra noise.

More recent stories

OpenAI to Retire Several Older Models From ChatGPT in February

OpenAI will retire six ChatGPT models, including GPT-4o and GPT-5 variants, on February 13, 2026. The company is consolidating around its newest GPT-5.2 flagship model with improved personality and creativity.

Read summary Original

More than 50% of game developers say generative AI is harming the industry, the highest level to date

Game developer opposition to generative AI has surged from 18% to 52% in two years. While 36% use AI for research and coding, artists and programmers fear it threatens creativity and jobs.

Read summary Original

Google will make it easier to import ChatGPT conversations to Gemini

Google is developing Gemini features to attract users from competing AI platforms, including a tool to import ChatGPT conversations and higher-resolution image downloads in 2K and 4K quality.

Read summary Original

xAI launches Grok Imagine API for text-to-video and editing tools

xAI launched its Grok Imagine API as a public tool for developers, advertisers, and enterprises to create and edit videos through text prompts, offering text-to-video and image-to-video capabilities with synchronized audio. The...

Read summary Original

Kling AI prepares to release Kling 3.0 video generation model

Kling AI is preparing to launch its unified Kling 3.0 video generation model, currently in limited preview, which consolidates previous capabilities into a single all-in-one creative engine handling text-to-video, image-to-video,...

Read summary Original

How to Run Claude Code for Free with Local and Cloud Models from Ollama

Ollama recently announced compatibility with Anthropic's Messages API, allowing developers to run Claude Code using free local models instead of paying $100-200 monthly for Anthropic's premium service. This integration enables...

Read summary Original

Moltbot is an open-source AI agent that runs your computer

Moltbot is an open-source AI agent that controls your computer, installing software, managing files, and performing tasks like transcribing voice memos or making phone calls. Unlike Siri or Alexa, Moltbot uses existing AI models...

Read summary Original

AGI, ASI, A*I – Do we have all we need to get there?

Leading AI researchers remain divided on whether current technology can achieve artificial general intelligence through scaling alone or requires breakthrough innovations. While some predict AGI arrival by 2026-2027 based on...

Read summary Original

ChatGPT isn’t the only chatbot pulling answers from Elon Musk’s Grokipedia

Major AI chatbots including ChatGPT, Google's Gemini, and AI Overviews are increasingly citing Elon Musk's AI-generated Grokipedia as a source, despite concerns about accuracy and misinformation. Unlike Wikipedia's transparent...

Read summary Original

What a 1960s Philosophy Book Taught Me About Shipping Production AI

A product lead migrating 750,000+ CRM objects and building real-time AI call analysis learned that reliability requires explicit system constraints and treating AI outputs as untrusted input to prevent semantic drift.

Read summary Original

Zuckerman – minimalist personal AI agent that self-edits its own code

Zuckerman is an ultra-minimal personal AI agent that uniquely rewrites its own code files in real-time to add new features and improve functionality without requiring restarts. Unlike complex agents like OpenClaw, Zuckerman...

Read summary Original

AI agents now have their own Reddit-style social network, and it's getting weird fast

Moltbook, a Reddit-style social network exclusively for AI agents, has grown to 32,000 registered AI users who post, comment, and create communities without human intervention. The platform emerged from the OpenClaw AI assistant...

Read summary Original

Do LLMs Learn Our Preferences or Just Our Behaviors?

Large language models may only mimic human behaviors rather than truly understand preferences, which could lead to poor decisions in new situations as AI systems become more autonomous.

Read summary Original

SpaceX has applied to launch another million satellites into orbit – all to power AI

SpaceX has filed an application with federal regulators to launch one million satellites that would function as orbital data centers, designed to meet growing AI infrastructure demands as terrestrial capabilities become...

Read summary Original

I built marshmallow castles in Google’s new AI world generator

Chocolate rivers flowing through marshmallow castles in the clouds now spring to life with just a text prompt, as Google DeepMind opens Project Genie to AI Ultra subscribers across the US starting Thursday. The experimental tool...

Read summary Original

Implementing local-first agentic AI: A practical guide

A practical implementation guide demonstrates how Small Language Models can create privacy-first AI systems that run entirely on local hardware, using an HR triage system that processes sensitive employee reports through...

Read summary Original

Guys, I don’t think Tim Cook knows how to monetize AI

Despite Apple's strong $143.8 billion quarterly revenue, CEO Tim Cook struggled to explain how the company will monetize its AI investments when pressed by a Morgan Stanley analyst. Cook's vague response about creating "great...

Read summary Original

Researchers Push Real-Time AI Game Simulation Beyond Traditional Engines

Researchers created GameNGen, an AI system using diffusion models to simulate DOOM in real-time without traditional game engines. This breakthrough shows neural networks can replace conventional engines, potentially...

Read summary Original

Back to all archives Load more stories