Observed Signal · Jun 30, 2025 · Analysis · Source: Trending Topics · Impact: 3/5 · Sentiment: Positive
Karpathy: LLMs Are the New Operating Systems
Andrej Karpathy, AI researcher and investor, outlined a vision in which large language models (LLMs) become universal infrastructure, comparable to electricity, operating systems, and semiconductors. He categorizes software evolution into three stages: Software 1.0 (traditional programming), Software 2.0 (neural networks), and Software 3.0 (LLMs where natural language becomes the programming interface). Karpathy draws parallels between GPU-based training companies and fabless semiconductor firms, comparing Google's in-house TPU training to Intel's manufacturing model, and citing xAI's Colossus cluster of 100,000 H100 GPUs as a scaling example. He argues LLMs function like new operating systems, where system and user prompts mirror kernel and user space, allowing applications such as Cursor to run across multiple models including GPT o3, Claude 4-Sonnet, Gemini 2.5-Pro, and DeepSeek.
Influential AI researcher presents a strategic framework positioning LLMs as universal infrastructure and operating systems, informing the future of AI-driven industries including advertising technology.
Track OpenAI Signals & Market Shifts in Real-Time
Polaris7 autonomous intelligence agents track regulatory filings, primary sources, executive changes, and deal flow 24/7. Create your free Explorer workspace to monitor these entities.
Key Takeaways & Evidence Grounding
- Karpathy categorizes software development into three stages: Software 1.0 (programming), 2.0 (neural networks), and 3.0 (LLMs).
- He compares LLMs to operating systems, drawing parallels between system/user prompts and kernel/user space.
- Karpathy likens NVIDIA-GPU-based training companies to fabless semiconductor firms and Google's TPU training to Intel.
- He cites xAI's Colossus cluster with 100,000 H100 GPUs as an example of massive scaling.
- He states LLM applications like Cursor can run across models such as GPT o3, Claude 4-Sonnet, Gemini 2.5-Pro, and DeepSeek.
Connected Companies & Entities
10 Entities mapped“Previously worked at OpenAI, then returned to work on GPT-4....”
“Moved to Tesla as head of AI....”
“Currently works as investor, including in Magic....”
“Google, which trains on its own TPUs, corresponds to companies like Intel that have their own manufacturing facilities....”
“Companies training on NVIDIA GPUs are compared to 'fabless' semiconductor companies....”
“Google corresponds to Intel, which has its own manufacturing facilities....”
“An LLM application like Cursor can run on GPT o3, Claude 4-Sonnet (Anthropic), Gemini 2.5-Pro, or DeepSeek....”
“LLM applications can run on GPT o3, Claude 4-Sonnet, Gemini 2.5-Pro, or DeepSeek....”
“An application like VS Code (Microsoft) can run on Windows 10/11, Mac, or Linux....”
Ontology Mapping & Concepts
Related Market Signals & Shifts
Recent verified developments and strategic activity across this market segment.
Fired OpenAI Researchers Dispute Misconduct Claims, Warn of Chilling Effect
Three OpenAI safety researchers, Jasmine Wang, Tomek Korbak, and Mikita Balesni, who were fired last week, have published an open letter denying allegations of mishandling sensitive information. They warn that their dismissal creates a chilling effect that stifles AI safety work and open dialogue within the company. The researchers say they acted in good faith and within company norms, and that the reasons for their firing are unclear. They urge OpenAI to uphold its commitments to third-party safety auditors and maintain a transparent culture. OpenAI responded with an internal memo denying retaliation and stating that employees are not terminated for raising concerns, but did not address specific policy violations or circumstances of the dismissal. The dispute highlights tensions between internal safety research and company communication policies at a time when OpenAI faces scrutiny over safety incidents.
Ben Affleck's AI Knowledge Goes Viral
Ben Affleck, the actor who sold his AI filmmaking startup to Netflix for a reported $587 million, has gone viral for demonstrating deep understanding of AI technology in recent interviews. He discussed machine learning, neural networks, and even admitted to writing Python scripts. Affleck detailed his journey from analog to digital film, his use of AI in visual effects, and his creation of a proprietary dataset for ethical AI filmmaking. He also visited OpenAI and used his celebrity status to access emerging tech. Affleck addressed AI's potential, expressing concerns about its effect on education and learned helplessness rather than existential threats. He used AI in his movie 'Animals' and emphasized the additive nature of AI in filmmaking.
AI Leaderboard Arena Raises $200M at $3.1B Valuation
Arena, the AI leaderboard platform that originated as a UC Berkeley research project, has raised a $200 million Series B round at a $3.1 billion valuation. The round was led by Lightspeed Venture Partners and Khosla Ventures, with participation from Salesforce Ventures, 01 Advisors, Dell Technologies Capital, Endeavor Catalyst, a16z, Felicis, and others. This follows the company's announcement in June that it reached $100 million in annualized run-rate revenue. Arena provides a crowdsourced platform where users rate AI model outputs, and it has introduced a commercial product called AI Evaluations to offer detailed performance analytics. The company has also added a new 'alignment' category to its leaderboard, ranking models on issues like unauthorized actions and deceptive completion. Arena's valuation has nearly doubled in about 10 months, from $1.7 billion post-money in January to $3.1 billion now.
Track Real-Time Market Signals & Shifts
Set up custom watchlists to receive automated, evidence-grounded executive digests whenever material signals or shifts occur across your tracked landscape.