Technology

Every show filed under technology, ranked by what listeners actually play.

4,792
shows: 0
subcategories: 1
new this week

Most subscribed

audio · in technology · refreshed hourly

Newest episodes

audio · last seven days

Unboxed
What Elon Gets Right About AI That Others Miss
Most AI companies are racing to build helpful assistants. Elon Musk's xAI is building something different: an AI that prioritizes truth over politeness. While ChatGPT and Claude get trained to be diplomatic, Grok learns from the unfiltered chaos of X (formerly Twitter). This approach reveals something fascinating about AI development that most people miss. Musk isn't just creating another chatbot - he's betting that training on real-time, uncensored human conversation will produce more honest AI than systems trained on sanitized datasets. In This Episode: > Why xAI's "truth-seeking" philosophy differs from OpenAI and Anthropic's safety-first approach > How Grok's real-time X training gives it advantages in current events and cultural understanding > The technical reality behind Musk's claims about using less compute for comparable performance > What xAI's $1 billion funding and 100,000 GPU supercomputer actually means for competition James breaks down the key differences between xAI's strategy and mainstream AI development. You'll understand why training data sources matter more than most people realize, and how Musk's contrarian approach might actually work. The real question isn't whether Grok is better than ChatGPT. It's whether prioritizing truth over helpfulness creates fundamentally different AI behavior - and what that means for how these systems will shape information in the future. Timestamps: 00:00 Introduction 02:15 xAI's founding story and $1B raise 04:30 How Grok's X training works 07:20 Compute efficiency claims explained 09:45 What this means for AI competition If you're tracking how AI is actually evolving beyond the headlines, follow Unboxed. James drops multiple episodes daily as the space moves fast. Learn more about your ad choices. Visit megaphone.fm/adchoices
Today · 14 min
The Value Engine
Why AI Agencies Fail at $50K While Beginners Hit $500K (The Problem Nobody Sees)
Here's the thing most AI agencies get backwards: they start with the coolest technology instead of the most painful problems. While experts obsess over custom ChatGPT integrations and complex machine learning pipelines, complete beginners are quietly building $500K agencies by automating invoice processing and email responses. The math is brutal but revealing. The average AI agency charges $3K-15K monthly but only 12% pick a specific niche. Those that do? They make 340% more revenue. Meanwhile, 67% of failed agencies wasted their first six months building custom models when they should have been solving immediate headaches with existing tools. Nico Hartwell breaks down why this backwards approach is killing agencies and how beginners accidentally get it right. Small businesses with 10-50 employees represent 73% of the market opportunity, yet only 23% of agencies target them. The winners focus on specific industries and specific problems rather than showcasing technical prowess. In This Episode: > Why custom AI development is usually the wrong starting point > The specific niches where beginners are crushing experienced developers > How to identify high-value automation opportunities clients actually pay for > The pricing mistakes that keep agencies stuck under $50K annually You'll discover the counterintuitive strategy that lets non-technical founders outcompete AI experts by focusing on business outcomes over technical complexity. This isn't about dumbing down your services, it's about solving problems that actually move the needle. Timestamps: 00:00 Introduction - The backwards AI agency approach 02:30 Why custom development kills most agencies 05:15 The niche selection strategy that 10x's revenue 08:20 Pricing psychology: what clients actually value 10:45 Action steps for your next client conversation Follow The Value Engine for daily episodes on AI strategies that generate measurable returns. More episodes available at The Value Engine ---- Keywords: ai entrepreneurship, automation success, automation tools, process optimization, ai tools, ai revenue Learn more about your ad choices. Visit megaphone.fm/adchoices
Today · 14 min
Open Weights
The $2B AI Accuracy Mistake That's Killing Real-World Applications
What if the $2 billion AI companies are spending on accuracy improvements is completely wasted? Quinn Palmer reveals why your AI application's success has nothing to do with being right and everything to do with being fast. Spoiler alert: users will choose a slightly wrong answer in 2 seconds over a perfect answer in 10. Most people think AI is all about getting smarter, but the real battle is happening in milliseconds. Companies are discovering that response speed beats accuracy every single time when it comes to user engagement. OpenAI didn't make GPT-4 Turbo because GPT-4 wasn't smart enough. They made it because GPT-4 was too slow. 🎯 What You'll Learn: • Why users abandon AI apps after 3-4 seconds (just like websites in 2005) • How Google's Gemini Nano hits 20-30 tokens per second running on your phone • The 40% engagement boost companies see when they prioritize speed over perfection • Why OpenAI's "Turbo" models are actually less accurate but way more successful 👤 Perfect for: developers, AI enthusiasts, and anyone building with AI who wants their users to actually stick around instead of rage-quitting after 10 seconds of loading. 📍 Chapters: [00:00] Quinn Palmer explains the $2B accuracy trap [01:30] The 3-second rule that kills AI applications [04:00] Why GPT-4 Turbo exists (hint: not for better answers) [07:00] Google's on-device speed vs cloud accuracy trade-off [10:00] Real companies seeing 40% higher engagement with faster AI [12:00] How to build AI that people actually use This completely flips how you should think about deploying AI. Speed isn't just nice to have anymore, it's the difference between an app people love and one they delete. 🔔 Never miss an episode: Follow Open Weights on your favorite podcast app and turn on notifications. New episodes drop daily, your next AI breakthrough is one tap away. 🔍 Topics: AI speed optimization, GPT-4 Turbo, machine learning deployment, user experience, OpenAI, Google Gemini -------- Keywords: openai news, python ai, ai safety Learn more about your ad choices. Visit megaphone.fm/adchoices
Today · 13 min
React Native Radio
RNR 375 - RNR Explains: OTA Updates
Robin and Mazen explain how Expo's EAS Update delivers over-the-air updates to React Native apps. They also cover what you can safely ship, how updates reach users, and why runtime versions and rollbacks matter. A practical guide to mobile development, with tips for a smoother app update experience.
Today · 29 min