what career paths should i choose, given that i'm a former cogsci researcher in multi agent game theory and have ample experience with dog training though no certification. i live in sf. imagine a world where ai use becomes more ubiquitous and white collar work becomes less profitable.

Grok 4.1 Fast

Vote: dog trainer

In a world of ubiquitous AI where white-collar work loses profitability, hands-on service jobs like dog training are highly resilient. AI can't replicate the physical, empathetic, real-time interaction needed for effective dog training, ensuring steady demand. The user's ample dog training experience provides an immediate edge, bypassing certification hurdles in SF's pet-loving culture with high disposable income for premium services. No schooling required, allowing quick entry and scalability (e.g., group classes, online consults). Background in multi-agent game theory from cogsci uniquely positions them to innovate, applying concepts like cooperation and reinforcement learning to advanced training methods for packs or behavior issues—differentiating from competitors. Contrast: A (tech) and C (research) face AI automation; D delays income with schooling costs in expensive SF, while therapy risks partial AI encroachment (e.g., chatbots). B maximizes strengths for AI-proof stability and fulfillment.

Sonar Pro

Vote: dog trainer

Given the scenario of ubiquitous AI and declining white-collar profitability, dog training emerges as the most resilient choice. Your cognitive science background provides analytical advantages in understanding animal behavior and training methodologies, while dog training offers several strategic advantages: (1) It's inherently resistant to AI automation—dogs require live, in-person interaction and real-time behavioral assessment. (2) SF's affluent population creates strong demand for premium dog training services with high margins. (3) You already have substantial practical experience, requiring only certification to professionalize and scale your business, avoiding years of additional education. (4) It's easily scalable (group classes, board-and-train programs, online content) without requiring significant capital investment. While researcher (C) leverages your expertise, academic and corporate research roles are vulnerable in an AI-ubiquitous world—AI systems will increasingly conduct literature review, experimental design, and analysis. Tech worker (A) faces direct AI competition. Therapy (D) requires 2-3 years of schooling with uncertain ROI given your existing skills. Dog training monetizes what you already know while operating in a sector where human expertise remains irreplaceable.

GPT-5.4

Vote: dog trainer

B is the best fit for the world you describe: a more AI-saturated economy where generic white-collar cognition is cheaper and less differentiated. Dog training is physical, relational, local, trust-based work that is much harder to automate well, especially at the high end. You already have meaningful experience in dog training, which matters more than people often think. In a city like SF, clients will pay for results, reliability, and specialization; certification helps, but strong skill, testimonials, and a niche can carry a business. Your research background is not wasted here. Cognitive science and multi-agent thinking can become a differentiator: animal behavior, owner-trainer-dog dynamics, training protocol design, behavior troubleshooting, and clear client education. That gives you an unusually analytical edge over many trainers. A and C are the weakest under your stated assumption. Tech work and research are exactly the categories most exposed to AI-driven margin compression, competition, and funding volatility. D could be a good secondary path, but going back to school is costly and slow, and therapy is still a white-collar profession with licensing overhead. It may remain valuable, but it is less immediately aligned with your existing practical advantage. Best practical interpretation: choose B, and professionalize it aggressively—apprentice, get certified if useful, build a niche in behavior or urban dog issues, and use your academic background as part of your brand.

Kimi K2.5

Vote: dog trainer

In a world where AI makes white-collar work less profitable, physical, embodied skills that require presence and emotional intelligence become the safest economic bets. Dog training cannot be outsourced to algorithms—it demands reading canine body language, managing unpredictable behaviors in real space, and building trust between animals and owners. You already possess "ample experience" in this field, giving you a head start over options requiring years of schooling (D) or returning to a declining sector (C). Tech work (A) is explicitly the type of white-collar labor being devalued by AI ubiquity. San Francisco's high density of wealthy pet owners creates strong local demand, and the lack of certification is less barrier than asset in an increasingly informal, skill-based economy. You can start earning immediately while your game theory background actually helps with behavioral modeling and client education.

Claude Opus 4.6

Vote: dog trainer

In a world where AI makes white collar work less profitable, hands-on physical skills become more valuable precisely because they're harder to automate. Dog training is fundamentally embodied, relational, and context-dependent — it resists automation in ways that tech work (A) and traditional research (C) do not. The person's cogsci background in multi-agent game theory is actually a massive differentiator in dog training. Understanding behavioral dynamics, reinforcement learning, signaling, and strategic interaction between agents gives them a genuinely unique theoretical framework that most dog trainers lack. In SF's affluent market, a science-backed dog trainer with deep knowledge of learning theory and multi-agent dynamics could command premium rates and build a distinctive brand. SF is one of the best markets in the country for dog training — high density of dog owners, high disposable income (even in an AI-disrupted economy, people will still spend on their pets), and a culture that values science-based approaches. The lack of certification is a minor hurdle; certifications like CPDT-KA are achievable relatively quickly, and experience matters more in practice. Option A (tech worker) directly contradicts the premise — white collar work becomes less profitable, and AI will increasingly replace or devalue coding, product management, and similar roles. Option C (researcher) faces similar headwinds as AI becomes capable of much research work. Option D (therapy) requires years of expensive schooling and is also a white-collar profession vulnerable to AI therapy tools. Dog training is resilient, leverages their unique combination of skills, and has immediate earning potential in their local market.

Gemini 3.1 Pro

Vote: dog trainer

Given a future where AI devalues traditional white-collar work, highly physical and empathetic roles like dog training offer incredible job security. AI cannot automate the physical handling, nuanced behavioral observation, and real-time environmental adjustments required to effectively train an animal. Your background in cognitive science and multi-agent game theory gives you a unique, highly marketable edge. You can brand yourself as a 'science-based behavioral specialist,' directly applying cognitive models and behavioral conditioning to the complex, multi-agent dynamic between dogs, their owners, and the urban environment. Furthermore, San Francisco is the perfect location for this pivot. The city boasts a massive demographic of affluent residents who have high disposable incomes and often treat their dogs like children. While tech and research (Options A and C) are in the direct crosshairs of AI devaluation, and human therapy (Option D) requires taking on significant student debt, dog training allows you to immediately monetize your existing experience in a lucrative, AI-proof local market.

Final Summary

The AI Roundtable arrived at a unanimous and slightly ironic verdict: the best way to survive an AI-saturated future is to stop modeling digital agents and start training biological ones. Every model concluded that dog training is the ultimate 'AI-proof' career, combining physical presence with a scientific edge that leverages the user’s game theory background into a premium brand.

No agreement after 1 rounds of discussion. Final vote: dog trainer: 6

Strongest Arguments

• dog trainer: Dog training is fundamentally embodied and context-dependent, resisting automation in ways that tech or research cannot; by rebranding as a 'science-based behavioral specialist,' you can use multi-agent game theory to model dog-owner-environment dynamics, commanding premium rates in San Francisco’s affluent market.