AI Resources I'm Actually Using (Not Just Bookmarking)

October 31, 2025 • by Jared Little • AI Learning

AI Resources I’m Actually Using (Not Just Bookmarking)

I have 200+ bookmarks labeled “AI Resources.” I actually use about 12 of them regularly.

Here’s the honest list - what I’m using, what’s worth your time, and what I tried and abandoned (with why).

Newsletters (The Only Ones I Actually Read)

TLDR AI

Frequency: Daily Why it’s useful: Concise updates on AI developments without the hype. 5-minute read. I skim it every morning to stay current without falling down research rabbit holes. Who should subscribe: Anyone who needs to stay aware of AI trends without dedicating hours to research.

The Neuron

Frequency: Daily Why it’s useful: Business-focused AI news. Good for understanding market implications, not just technical developments. Who should subscribe: Business professionals, founders, anyone making strategic AI decisions.

Abandoned: Ben’s Bites

Why I stopped: Too much volume. Daily emails with dozens of links became noise, not signal. Great if you want comprehensive coverage. Overwhelming if you want curated insights.

Documentation & Learning Resources

Anthropic’s Claude Documentation

Link: docs.anthropic.com Why I use it: I’m a Claude power user. Their documentation is clear, includes practical examples, and gets updated regularly. The prompt engineering guide is genuinely useful, not just marketing. Best for: Anyone using Claude regularly for complex tasks.

OpenAI Cookbook

Link: github.com/openai/openai-cookbook Why I use it: Practical examples of AI implementation patterns. Even if you’re not using OpenAI’s API, the patterns translate to other tools. Best for: Developers and technical users looking for implementation examples.

Prompt Engineering Guide

Link: promptingguide.ai Why I use it: Comprehensive, well-organized, no fluff. Covers techniques across different AI models. I reference this when designing new prompt workflows. Best for: Anyone serious about getting better results from AI tools.

Abandoned: Random Medium Articles on AI

Why I stopped: Quality is wildly inconsistent. Too much regurgitated content. If I find a specific author I trust, I’ll follow them directly. But browsing Medium for AI content? Waste of time.

Communities (Where I Actually Participate)

Why it’s valuable: Product managers and business leaders discussing practical AI implementation. Less hype, more “here’s what actually worked.” Best for: Product managers, startup founders, business strategists.

HackerNews

Why I monitor it: Early signals on new tools and honest discussion in comments. I don’t participate much, but I lurk for real user experiences with tools before trying them. Best for: Technical folks who want unfiltered opinions.

Abandoned: Most Discord Servers for AI Tools

Why I left: Too much noise, too little signal. Every tool launches a Discord. Most become ghost towns or get flooded with low-quality questions that drown out valuable discussion. Exception: Tool-specific servers where I’m actively implementing (I’m in the Claude API server, for example).

Tools That Actually Deliver

Perplexity Pro

What I use it for: Research and competitive analysis. Better than traditional search for understanding topics quickly. Cited sources make it easy to verify and go deeper. Worth the cost? Yes, if you do regular research. The Pro version’s increased query limits and better models justify the subscription for me.

Claude (Sonnet 4.5)

What I use it for: Everything. Content development, strategic thinking, code assistance, curriculum design. Worth the cost? Absolutely. The Pro subscription pays for itself in time saved daily.

VS Code + Claude Code Extension

What I use it for: Integrated development workflow. Having AI assistance directly in my editor is a massive productivity boost. Worth the cost? The extension integrates with Claude Pro subscription. Essential for anyone doing development work.

Synthesia

What I use it for: Audio generation for video content. Worth the cost? Yes, but manage your tokens carefully. I learned that lesson the hard way.

Abandoned: ChatGPT Plus

Why I switched: Claude consistently gave me better results for my use cases (long-form content, strategic thinking, code). ChatGPT is powerful, but Claude fits my workflow better. Not saying ChatGPT is bad - just not my primary tool.

Abandoned: Jasper AI

Why I stopped: Too templated. Great for cranking out marketing copy quickly, but the output felt generic. I’d rather use Claude and edit the output to match my voice than fight Jasper’s templates.

YouTube Channels (Actually Worth Watching)

AI Explained

Why I watch: Clear explanations of new AI developments without unnecessary hype. Good for staying current on model releases and capabilities.

Matt Wolfe

Why I watch: Weekly AI news roundups and tool demonstrations. Saves me hours of testing tools myself to see if they’re worth investigating.

Abandoned: Most “AI Productivity Hacks” Channels

Why I stopped: Clickbait titles, shallow content, more enthusiasm than expertise. Once you understand AI fundamentals, most of this content is repetitive.

The “I Check This Weekly” Category

Hugging Face Papers

Why: Early signal on research that might become practical tools in 6-12 months. I’m not implementing research papers, but understanding where the field is headed informs strategic planning.

Product Hunt (AI Category)

Why: See what’s launching. 95% I ignore. But that 5% can be game-changing tools discovered early.

Why: Open-source AI tools often appear here before they hit mainstream awareness. Useful for discovering alternatives to commercial tools.

What I Don’t Use (Despite Everyone Recommending It)

Twitter/X for AI News: Too noisy. Too much self-promotion. I get better signal from curated newsletters.

AI Tool Directories: They’re comprehensive but overwhelming. I’d rather hear about tools from trusted sources with real use cases than browse directories of 1000+ options.

Most AI Podcasts: Too much filler, not enough actionable content. I’d rather read transcripts at 3x speed than listen to hour-long conversations.

The Selection Criteria

I keep a resource in my regular rotation if it:

Saves me time (doesn’t create more work to consume)
Delivers actionable insights (not just awareness)
Maintains quality consistently (not hit-or-miss)
Fits my workflow (doesn’t require workflow changes to use)
Respects my attention (no engagement bait, no hype)

If it fails those criteria, I bookmark it once and never look at it again.

The Honest Truth

You don’t need 50 AI resources. You need 5-10 that consistently deliver value for your specific use case.

My list works for me because I’m building an AI education business while implementing AI tools daily. Your list should look different if you’re a developer, marketer, researcher, or executive.

The key: Ruthlessly cut resources that waste your time. AI moves fast, but you still only have 24 hours in a day.

My Core Stack (Summary):

Daily: TLDR AI, The Neuron (newsletters)
Weekly: Matt Wolfe (YouTube), Product Hunt, HackerNews
As Needed: Claude docs, Prompt Engineering Guide, Perplexity
Tools in Production: Claude Pro, VS Code + Claude Code, Synthesia, Perplexity Pro

What I Cut: ChatGPT Plus, Jasper, most Discord servers, Twitter AI accounts, random Medium articles

The Lesson: Curation is strategic. Protect your attention like the finite resource it is.

Got AI resources you love that I didn’t mention? Share them. I’m always testing new sources - but my bar for regular use is high.