Andreessen believes artificial general intelligence — AI as capable as a human — was crossed approximately three months before this recording, with the latest model versions from OpenAI, Anthropic, and Google.
Where this was said
At 2:25:26 · chapter starts 2:25:00
This is the intellectual core of the AI section. Andreessen states plainly that he believes artificial general intelligence was achieved roughly three months before this recording, with GPT-5.5, Claude 4.6, Gemini 3.0, and Grok 4.3. [1] — Marc Andreessen "AGI crossed ~3 months ago per Andreessen: Andreessen believes artificial general intelligence — AI as capable as a human — was crossed appr…" 2:25:26 The reason this hasn't been treated as civilizational news: most people are still using models from last year and don't realize how different the latest versions are. He walks through the Turing test — 60 years as the gold standard for machine intelligence, never passed — and explains how AI blew through it during Christmas 2022 when ChatGPT launched, so decisively that the formal test was barely administered. It turned out intelligence is 'not that hard' once the key conceptual pieces — neural networks, transformers, gradient descent, reinforcement learning — were in place. Andrej Karpathy built a working LLM in 300 lines of code. People have ported LLMs to Texas Instrument calculators. Andreessen then explains training (building the model from scraped data) and inference (the model generating answers) as the two activities that fill those massive data centers everyone is arguing about.
Andreessen says that with the leading-edge AI models, 99% of answers he receives are better than what he would get from almost any human expert, even with his extensive access to top professionals.
For 60 years, the Turing test was the gold standard: if a human can't tell whether they're talking to a machine or a person, the machine wins. AI blew right through it during Christmas 2022 when ChatGPT launched — so fast and so decisively that nobody even bothered to formally run the test. It was yesterday's news before the field had time to celebrate.
AI blew through the Turing test over the Christmas 2022 holiday when ChatGPT launched, passing so decisively that the formal test was barely even administered.
AI researcher Andrej Karpathy has built a working large language model in just 300 lines of software code, demonstrating that foundational AI is simpler than widely assumed.
Ad-based monetization works well for game apps where users spend extended time in-session, as seen with Grid and Wordle.
Tool-focused apps like PuffCount are poor candidates for ad monetization because users don't stay in-session long enough.
A hard paywall is a screen that blocks all app features unless the user pays or starts a free trial — it cannot be dismissed.
Mobile apps are primarily monetized through either ads (best for games) or in-app purchases/subscriptions (best for tools).
According to the episode, YouTube outperforms every other social platform for building trust and driving SaaS conversions.
Vasco stated that the majority of his app's user base came directly from his YouTube channel.
SEO Bot features a 'Boost My Domain Rating' button that routes users directly to Listing Bot, an example of in-product cross-selling.
The founder's entire product portfolio is AI-related, making it easier to package products attractively for directories.
The founder attached their SaaS demo to the trending debate about whether AI coding is actually good enough to build a full SaaS product.
We use essential and analytics cookies to run Vuci. To understand how the site is used: Privacy Policy.
Install Vuci on your phone
Add it to your home screen for a faster, app-like experience.
Install Vuci on your phone
Tap the Share button, then “Add to Home Screen”.
A new version is available
Reload to get the latest Vuci.