"AI Brendans" or "Virtual Brendans" (2025)

10 hours ago

Brendan Gregg identifies two types of AI agents: AI Brendan (pattern matching for flame graphs/eBPF) and Virtual Brendan (trained on his publications but limited to ~15% of his job).
Commercial AI autotuners face pricing challenges, as customers may pay for one instance and copy tuning changes across fleets.
Secret tuning changes are impractical to hide; engineers can easily uncover them, and such secrecy risks blame during outages.
Published works are an incomplete and dated snapshot, missing half of Gregg's experience and real-time industry changes.
Historical precedent includes 'Virtual Adrian' (1994) and Intel's $650M Granulate acquisition, which was later shut down.
In-house or open-source AI tools may be more viable than commercial ones, avoiding secrecy and upstreaming issues.
AI agents could help address global performance engineering gaps and save energy, but should not replace human expertise entirely.

Startups older than two years likely have obsolete business plans due to rapid changes in AI and markets.
Venture capital has shifted heavily toward AI, making it harder for non-AI startups to raise funds.
AI development tools enable building MVPs in days, reducing the need for large engineering teams and challenging old assumptions about cost and speed.

The author delays building a local AI inference rig due to fast market changes and high hardware prices.
China may restrict export of top AI models, including open-weight ones like Qwen and GLM-5.2, threatening Western access.
Open-weight models from China enable affordable local and serverless AI; a ban could limit this access.

AI-related security incidents are common, as many enterprise AI systems were implemented hastily without adequate oversight.
Quantum computing and AI are being explored to address challenges in fusion energy, specifically in simulating tritium production.
Software engineers continue to command high salaries, especially at fast-growing companies, despite AI's complex impact on the tech industry.

AI agents cannot truly program; they are statistical models that produce increasingly hard-to-detect broken code.
The author experimented with agents for 6 months but found manual work more efficient; agents frontload progress but fail at polish.
AI is useful for prototypes and better than Google for searches but is not a substitute for software engineers.

The Department of Energy, Cleveland Clinic, and IBM are simulating molten salts to address the fusion fuel dilemma, focusing on tritium production.
Software engineers can earn high salaries at fast-growing companies, but AI's impact on the tech business is complex and varied.
Microsoft is restructuring its commercial and Xbox teams, acknowledging rapid global changes and the need to avoid complacency.

Startups older than two years likely have obsolete business plans due to rapid changes in AI and markets.
Venture capital has shifted heavily toward AI, making it harder for non-AI startups to raise funds.
AI development tools enable building MVPs in days, reducing the need for large engineering teams and challenging old assumptions about cost and speed.

The author delays building a local AI inference rig due to fast market changes and high hardware prices.
China may restrict export of top AI models, including open-weight ones like Qwen and GLM-5.2, threatening Western access.
Open-weight models from China enable affordable local and serverless AI; a ban could limit this access.

AI-related security incidents are common, as many enterprise AI systems were implemented hastily without adequate oversight.
Quantum computing and AI are being explored to address challenges in fusion energy, specifically in simulating tritium production.
Software engineers continue to command high salaries, especially at fast-growing companies, despite AI's complex impact on the tech industry.

AI agents cannot truly program; they are statistical models that produce increasingly hard-to-detect broken code.
The author experimented with agents for 6 months but found manual work more efficient; agents frontload progress but fail at polish.
AI is useful for prototypes and better than Google for searches but is not a substitute for software engineers.

The Department of Energy, Cleveland Clinic, and IBM are simulating molten salts to address the fusion fuel dilemma, focusing on tritium production.
Software engineers can earn high salaries at fast-growing companies, but AI's impact on the tech business is complex and varied.
Microsoft is restructuring its commercial and Xbox teams, acknowledging rapid global changes and the need to avoid complacency.

Hasty Briefsbeta