• Lore Brief
  • Posts
  • GPT-5 August Launch Rumors and the $100M Revenue Race

GPT-5 August Launch Rumors and the $100M Revenue Race

Plus: Microsoft raids DeepMind talent, Meta's neural wrist tech, and AI video generation goes commercial

Welcome to Lore Brief, your weekly edge in the age of AI.

This issue is brought to you by Factory, an engineer in every tab.

Factory is a Sequoia-backed startup that has created what I believe is the best agentic coding platform available. It’s better than Devin in all the tests I’ve done.

Whether you're prototyping quickly or building enterprise-grade solutions, Factory accelerates development. Major companies like Zapier, Bayer, MongoDB, and Framer utilize Factory to develop faster and resolve more bugs, allowing engineers to focus on what they enjoy doing, rather than debugging.

67 K-view demo: watch me and founder Matan Grinberg clone the core DocuSign functionality in 15 minutes!

Try Factory at Factory.ai and follow Factory on Twitter/X for more updates!

GPT-5 Rumors Heat Up

Speculation about OpenAI's GPT-5 release is ramping up, with industry chatter suggesting an early August launch after earlier July expectations fizzled. Sam Altman fueled the hype during a podcast appearance, describing the model as surpassing human intelligence in most areas and sharing how it solved a puzzle that stumped him.

  • Tech reporter rumors suggest GPT-5 will come with smaller "mini" and "nano" versions optimized for different use cases

  • Altman's appearance on Theo Von's podcast included stories about being genuinely surprised by the model's capabilities

  • Industry sources point to delays from the original July timeline, with estimates ranging from late July through September

The mounting anticipation suggests something significant is coming. Whether August delivers remains to be seen.

Lovable's Revenue Claims

AI startup Lovable claims it reached $100 million in annual recurring revenue within eight months - faster growth than OpenAI's early trajectory - while launching an improved coding agent that significantly reduces errors. Their latest update enables AI to think iteratively, search the web, and debug code like an experienced developer.

  • Founder Anton Osika announced the milestone on X, emphasizing advantages over traditional linear AI workflows for complex tasks

  • The enhanced agent can read files, analyze debug logs, and continuously refine its approach until tasks are completed correctly

  • Social media reactions are mixed, with supporters praising the achievement and skeptics questioning the revenue figures without verified proof

  • Defenders argue the growth reflects genuine success from a small team democratizing app development for non-programmers

If accurate, these numbers position Lovable as a major player in AI-assisted development tools.

GitHub's Spark announced

Microsoft unveiled GitHub Spark, a Copilot-powered feature that builds complete applications from plain English descriptions. The tool aims to democratize software development by letting anyone transform ideas into functional apps without coding knowledge.

  • The AI tool handles full-stack development, managing both frontend and backend components based on conversational instructions

  • Reception highlights excitement from beginners, though questions remain about handling complex application requirements

Spark could significantly lower barriers to app development for non-technical users.

The Next Wave: Interview with Demis Hassabis, CEO of Google DeepMind

Meta's Neural Interface

Meta revealed wrist-worn technology that interprets muscle signals to control devices through subtle hand gestures. The system translates neural intentions into digital actions for more intuitive, hands-free interaction.

  • Demonstration videos show precise control over tasks like scrolling and selection without touching screens

  • Meta's announcement thread details how the system combines multiple signal types for accurate gesture recognition

  • The technology builds on previous research but advances practical implementation of intent-based computing

The work offers a glimpse into potential future interface paradigms beyond traditional touch and voice control.

AI Talent Wars Intensify

Competition for top AI researchers has escalated dramatically, with Microsoft recruiting 24 scientists from Google DeepMind in one day, followed by Meta hiring three key researchers who contributed to DeepMind's recent mathematical breakthrough.

  • The aggressive recruiting demonstrates how major tech companies are raiding competitors for expertise in cutting-edge AI research

  • Industry observers note that while DeepMind retains strong leadership, losing experienced researchers impacts ongoing projects

  • The hiring frenzy underscores the current scarcity of top-tier AI talent across the industry

Veo 3's Commercial Potential

Google's Veo 3 video generation model created a polished IKEA advertisement from a single JSON prompt, producing furniture assembly footage that achieved viral attention for its professional quality and smooth execution.

  • Designer Salma Aboukar shared the test results, showcasing detailed lighting, motion, and product presentation that rivals traditional advertising production

  • The prompt specified scene progression step-by-step, requiring no additional refinement or direction

  • Other users have recreated similar versions, praising the model's ability to capture specific brand aesthetics and product demonstration styles

  • The result demonstrates Veo 3's potential for brands seeking rapid, cost-effective commercial video content

This capability could significantly streamline advertising production workflows and reduce costs for marketing teams.

HeyGen's Avatar IV

HeyGen demonstrated their new Avatar IV model, which transforms a single photo and audio file into a dynamic singing avatar with natural movement and style. The technology represents a significant leap in creating engaging video content without expensive production setups.

  • CEO Joshua Xu shared demonstration clips highlighting the system's ability to capture authentic performance energy

  • Early reactions praise the potential for instant music videos and advertising content

  • Users note some synchronization issues remain, but the core technology appears solid

HeyGen continues advancing photorealistic avatar technology that could reshape rapid content creation.

That’s it for today.

Consider forwarding Lore Brief to a colleague to help them get ahead in the AI Age.

-Nathan Lands
ConnectX | LinkedIn
Listen to The Next WaveApple | Spotify | YouTube

(Disclosure: I may own equity in companies mentioned in Lore Brief.)