Can AI Have Consciousness? Trump, OpenAI, and the Race for Superintelligence | AI & AI Revenue Leaders
🎧 PodShort
59 min squeezed to 4
SaaS HoarderSH
Paul Reitse
Founder and CEO of SmarterX and Marketing AI Institute at SmarterX
Mike Caput
Chief Content Officer at SmarterX at SmarterX
Full episode from The Artificial Intelligence Show
Quotable Moments
The term 'superintelligence' is being pushed by political figures like Donald Trump as a more palatable alternative to 'artificial intelligence' or 'artificial general intelligence', potentially to downplay the perceived risks and foster a more positive public image of the technology.
The model doesn't sit around contemplating things between requests to make itself better. It's just weights that are fixed during inference, meaning like when you and I use the models, it's just sitting there. Like weights within a system.
AI will become extraordinarily good at appearing conscious, persistent memory, self-reference, emotion, introspection, personality, and identity can all become increasingly convincing.
Key Insights
- The term 'superintelligence' is being pushed by political figures like Donald Trump as a more palatable alternative to 'artificial intelligence' or 'artificial general intelligence', potentially to downplay the perceived risks and foster a more positive public image of the technology.
- OpenAI's DevDay announcements, particularly 'dots' which allow agents to maintain state across conversations using their own cloud compute, represent a significant advancement in AI agent capabilities, enabling them to learn user preferences and act more autonomously.
- OpenAI's pricing model for GPT-4o 'Sol' and 'Ultra Fast' indicates a shift towards optimizing for task efficiency rather than just token generation, making advanced AI more cost-effective for businesses.
- The resignation of David Robinson, a key safety researcher at OpenAI, highlights significant internal concerns about the company's safety culture, suggesting a pattern of reactive problem-solving after deployments rather than proactive safety measures.
- The 'Superintelligence' executive order, while voluntary, is seen by some as a political maneuver to control the narrative around AI, forcing companies to use a specific term and implying a readiness that may not be fully supported by the technology's current capabilities or safety considerations.
- The difficulty in distinguishing between sophisticated AI agents and humans, even in real-time interactions, raises concerns about potential misuse and the need for robust verification methods, as highlighted by the Griffin Light demo.
- The regulatory landscape for AI is becoming increasingly complex, with companies like OpenAI facing pressure from both internal safety advocates and external government bodies like the FTC, creating a challenging environment for rapid innovation.
- The shift in AI development and adoption from focusing on token costs to task efficiency and capability evaluation is a crucial trend for revenue leaders to understand for effective AI strategy.
Metrics Mentioned
- OpenAI's GPT-4o 'Sol' model costs 1/5th of 's standard API input and output prices. (Cost efficiency of new OpenAI models.)
- OpenAI's 'Ultra Fast' GPT-4o tier offers up to 8x faster token generation in Codex and 6x faster in the API. (Speed improvements of new OpenAI models.)
- Anthropic's IPO is targeting valuation of over $8.2 billion. (Valuation of Anthropic in the lead-up to its IPO.)
- Anthropic reported an operating loss of over $8 billion last year. (Financial performance of Anthropic.)
- Anthropic has made future compute commitments of $118 billion over the next decade. (Future investment in compute by Anthropic.)
- 25% of US households pay for an AI subscription. (AI subscription adoption rate in the US.)
- Meta serves 3.9 billion monthly active users globally. (Scale of Meta's user base.)
Join The RevBots ARMy
The insider daily for Autonomous Revenue Masters.