While what follows is part education and part entertainment, a few grounding principles matter. First: the right stance is moderate optimism—not the techno-utopian “abundance” narrative, and not the doom-loop “end of the world” fatalism. Second: in markets, the bias still leans toward risk assets over time—not because things are calm, but because debasement keeps winning over deflation, even as volatility rises.

With that in mind, let’s zoom out.

The world is shifting beneath our feet. Across economics, politics, technology, and culture, the assumptions that defined the post–Cold War order are fraying—sometimes quietly, sometimes all at once. We are living through what historian Neil Howe might call the climax phase of a Fourth Turning, what Peter Turchin describes as elite overproduction and structural instability, and what Grant Williams and Demetri Kofinas have now branded The One Hundred Year Pivot: a rare civilizational inflection point in which the global architecture of power, economics, and social cohesion is being rebuilt in real time.

This month’s collection of links highlights some of the most important voices trying to map this transition. From the breakdown of political trust and the rise of “passenger-seat societies,” to mounting intergenerational and regional fractures within countries like Canada, to the emergence of new geopolitical and financial blocs, the throughline is clear: the old frameworks can no longer explain the world we’re actually living in. Whether we are investors, policymakers, or simply citizens trying to prepare our families for what comes next, new mental models are no longer optional—they are survival tools.

In the selections below, you’ll find conversations and essays that explore:

• The collapse of institutional legitimacy and social trust
• The rise of populist and elite-aspirant movements like the credentialed precariat
• The shift from globalized efficiency toward nationalized resilience and security
• The internal fractures—regional, generational, public vs private—that threaten stable democracies
• The early architecture of a new investment paradigm emerging out of chaos

These aren’t just intellectual curiosities. They are signals of a world reordering itself. The goal of this roundup is not to prescribe a single narrative, but to help you see the moving parts: the cultural mood shifts, the economic mechanisms, the social ruptures, the demographic fault lines, and the geopolitical incentives that together shape what history may remember as the 100-Year Pivot.

Hundred Year Pivot podcast series

Demetri Kofinas and Grant Williams have teamed up to launch “The One Hundred Year Pivot,” a new podcast series that explores the once-in-a-century economic, political, and geopolitical reordering currently underway. These transformational changes demand that we develop new maps and investment frameworks to help us navigate the challenges ahead.
Over the course of the series, Grant and Demetri will be speaking with the smartest and most plugged-in people they know to help us chart a path through this gathering storm and position ourselves, our organizations, our families, and our portfolios for the changes to come.
You are invited to join them on this journey as they explore the implications of:
-The breakdown of existing political, monetary, and social institutions
-The emergence of new economic and financial trading blocks
-The reprioritization of national and economic security
-The channeling of national savings into a new military and technological arms race
-The development of a new investment paradigm with historic implications for investors

Some specific episodes to highlight:

• https://www.youtube.com/watch?v=nnIv60_Y3jc – What Happens When Social Trust Collapses? | Peter Atwater (highlighted below)
• https://ttmygh.podbean.com/e/the-hundred-year-pivot-ep-2-neil-howe/ Neil Howe
• https://ttmygh.podbean.com/e/the-hundred-year-pivot-ep-1-demetri-kofinas/ Episode 1
• https://ttmygh.podbean.com/e/the-grant-williams-podcast-ep-90-demetri-kofinas/
• https://www.grant-williams.com/2025/fin-de-siecle/ Grant Williams Podcast article

What Happens When Social Trust Collapses? | Peter Atwater – Hundred Year Pivot podcast series https://www.youtube.com/watch?v=nnIv60_Y3jc

Episode 447 is the tenth installment in the Hundred Year Pivot podcast series. In it, Demetri Kofinas and Grant Williams speak with Peter Atwater, a recognized expert on the impact of confidence and mood on individual and group decision-making. Peter’s work has been instrumental not only in helping his clients, students, and readers understand how people make decisions under conditions of chaos and uncertainty, but also in showing how to use those insights to gain a competitive advantage.

The three begin their conversation by mapping out Peter’s Confidence Quadrant and using it to explain how individuals and crowds move between feelings of high and low certainty and control. They explore the immense contemporary pressure people feel to conform—particularly on social media and in an educational system that rewards conformity. They discuss financial nihilism, the search for community in unusual places, and why real, in-person networks, relationships, and conversations matter now more than ever.

The second half of the episode turns to the political economy of confidence: the rise of leader-centric movements, the fragility of “passenger-seat” societies, the lasting impact of COVID, and the “velvet-rope” economy that has transformed everything from airline cabins to financing. They discuss wealth stratification, powerlessness, the risks of a genuine socialist revolution, the search .

Paikin on new book, Breaking Point: The New Big Shifts Putting Canada at Risk, – https://www.youtube.com/watch?v=7WJ0y7s_BZ8 

Canada US relations have never been this bad, you say. But believe it or not, Trump is not the biggest threat our country faces. our own worst enemy is ourselves. Now, we’re of course going to dive in deep and explore this statement from many  different angles, but John, just start us off with the thesis. Why do you think we’re a bigger problem to ourselves than Trump is? Well, we’ve always said that Canada’s the one of the most successful countries in the world. I’ve always said that, but we’ve always known as well that there are things um tend to keep getting picked down the road and we keep thinking, well, we can manage this, we can manage that. Uh but some of these um let’s mix our metaphors. Chickens are uh coming home to roost and they involve both what we call horizontal and vertical divisions. So um we have uh a western Canada that at least in the prairie provinces is more alienated than at any time in the country’s history at the very same time that it looks as though there’s going to be another separatist government in Quebec that is promising a referendum. um and the the federal government keeps intruding in areas of provincial jurisdiction and it keeps creating more and more uh flash points. So there’s that division and then there’s another division which we can get into a bit later which is the intergenerational division um the a group of young people millennials and gen zeds who rightly feel uh that people uh of our generation have deprived them of much of the opportunity uh that we had when we were their age. So between the between the the generational tensions and the regional tensions and and issues that we have simply not properly addressed from immigration to defense uh to the media, we think the country is um at an inflection point and perhaps at a breaking point…..

Another potential division is private vs public sector due to AI pressures – “I surmise that a massive dislocation in private sector employment due to the introduction of AI technology predominantly, but not exclusively, in the service sector entirely explained by the unavoidable need to increase productivity, will lead to such a screaming disparity in the risk-reward profile of public and private sector jobs that it will become socially unacceptable to maintain the status quo” via https://www.grant-williams.com/2025/fin-de-siecle/

Peter Turchin – author of End Times – Revolt of the Credentialed Precariat https://open.substack.com/pub/peterturchin/p/revolt-of-the-credentialed-precariat?utm_source=share&utm_medium=android&r=1d8cv –

The main political news of the week is Zohran Mamdani winning New York City’s mayoral race. Quite a number of people, who’ve read End Times, commented that the Mamdani Moment is a perfect illustration of the “credentialed precariat” (see also this comment on X). 

For the background on credentialed precariat, here’s what I wrote in Chapter 4 of End Times: “Guy Standing, who injected the term ‘precariat’ into the public consciousness, sees degree holders as one of the precariat factions. This group

‘consists of people who go to college, promised by their parents, teachers and politicians that this will grant them a career. They soon realize they were sold a lottery ticket and come out without a future and with plenty of debt. This faction is dangerous in a more positive way. They are unlikely to support populists. But they also reject old conservative or social democratic political parties. Intuitively, they are looking for a new politics of paradise, which they do not see in the old political spectrum or in such bodies as trade unions (Meet the precariat, the new global class fuelling the rise of populism).’

History (and CrisisDB) tells us that the credentialed precariat or, in the jargon of Cliodynamics, frustrated elite aspirants, is the most dangerous class for societal stability.” (end of quote) 

My main interest is what this development may tell us about the evolution of political parties in America……

Fin de Siecle article – Grant Williams – https://www.grant-williams.com/2025/fin-de-siecle/ Grant Williams Podcast article

Like its monetary system, Western society has always had its own moral anchor—the Judeo-Christian tradition which has profoundly shaped the foundations of Western society, influencing its values, institutions, and cultural achievements. Central to this tradition is the belief in the inherent dignity and equality of all individuals, rooted in the biblical concept of being human, created in the image of God…..Just as the financial system will eventually be forced to return to its anchor—gold—to staunch the bloodletting, society will find itself in a similar position and a return to a spiritual/faith-based order feels inevitable to me. These twin reversions will form the basis of the First Turning. We can but hope they arrive sooner rather than later”

In late 2023 he saw the potential for both disruption and productivity being offered by the emergence of AI, notably ChatGPT, Claude and other early providers, and directed his tech team to begin building a specialised content writing tool for digital marketing. Within six months he was telling me that the quality of content production from the algorithm, given their ability to write highly specific prompts, was matching the creative output of his best  writers and within nine months, was able to create entire strategy briefs, extracted from deep subject matter and industry research. In combination with other AI-based process tools, the technology can now identify innovative ideal customer typologies, simulate accurate target customer group responses, draft individualised campaign emails, extract the emails, pen LinkedIn posts, create and illustrate in-depth white papers, blog posts and other content and then execute the campaign, whilst constantly monitoring the feedback and results, as well as adjusting the campaign in a constant feedback loop to optimise the responses. All this can be designed and executed within an hour rather than the weeks and months a well-staffed agency would have needed to have expended in time and resources not even a year ago.

However, I surmise that there is much greater explosive potential in an altogether different state of inequality within civil society—one which could become untenable in the evolving AI revolution. Namely the disparity between the public and private sector workforces. The main bone of contention is—and in my estimation will certainly be in future—the almost complete absence of any pressure to increase productivity or service quality which is referred to obliquely in the section ‘Working Conditions’ as ‘less stressful due to fewer competitive pressures.’ I surmise that a massive dislocation in private sector employment due to the introduction of AI technology predominantly, but not exclusively, in the service sector entirely explained by the unavoidable need to increase productivity, will lead to such a screaming disparity in the risk-reward profile of public and private sector jobs that it will become socially unacceptable to maintain the status quo.

Like a chronic illness, we often attempt to treat only the most visible symptoms, failing to address the root cause—and in doing so, we often worsen the condition. In medicine, merely suppressing symptoms can accelerate  deterioration; for example, managing lupus-related joint pain with painkillers does nothing to halt the underlying  autoimmune disorder, which continues to damage the body. Similarly, modern societies apply superficial solutions—monetary stimulus, welfare expansion, regulatory interventions—that provide short-term relief but fail to resolve systemic dysfunction. More often than not, these very interventions deepen the problems they aim to fix.

Throughout history, mankind has longed for strong leaders to restore order and heal the ailments of their time. Julius Caesar exploited economic inequality and political corruption, presenting himself as a champion of the people against the entrenched senatorial elite—only to become a dictator. Napoleon Bonaparte promised order, stability, and national glory to a France exhausted by revolution and chaos—yet he ultimately led the nation into ruin. History offers countless such examples. And yet, today, the pace of our degeneration appears to have accelerated at unprecedented speed. Why?

The debasement of money over the last thirty years has certainly eroded our social and economic fabric in  profound ways. But while this may explain financial turmoil, it does not fully account for the cascading crises that now touch every facet of life. What, then, is the deeper disease at the heart of our time?

At its core, I argue that the crises we face today are the culmination of a long historical process—one where the erosion of spiritual and moral foundations has led to state overreach, economic instability, and ultimately, the rise of authoritarianism.

The roots of decline

My own exploration to find the key element of all these symptoms has led me to reflect on their root causes. Here is a bird’s-eye view of such reflections.

The modern world has undergone a profound philosophical shift, one marked by the erosion of individual responsibility and the state’s increasing role as provider. It began in the 17th–18th century period of the Enlightenment, the Age of Reason, with its emphasis on rationality, science, and humanism. Rational governance excluded God as the central authority in political life. Reason became paramount.

….

Finally, in the age of consumerism and technocracy, people placed their faith in economic and technological progress, often mediated through State institutions. The State became not only a regulator and provider but also central to modern identity and life, replacing God as the ultimate guarantor of order and well-being.

In the end, the denial of scarcity creates a fragile economy that is over-leveraged and vulnerable to shocks. 

Misallocation of resources exacerbates inequality and undermines trust in institutions. The emboldening of a speculative culture diverts resources from productive enterprises to financial alchemy and manipulation, eroding the foundations of real economic growth.

Fragility, and the attendant rise in state power, in my view, also ultimately erode individual freedom, property rights, freedom of expression, and the rule of law.

The erosion of the rule of law

The rule of law is the cornerstone of any free society, ensuring that all individuals, regardless of their status or position, are subject to the same laws and protections. However, as institutions weaken, selective enforcement of laws becomes more common. This creates an environment where the elite or those in power can escape legal consequences, leading to a sense of injustice and disenfranchisement among the general populace.

Corruption becomes a major issue, and as it spreads, laws become tools for maintaining power rather than mechanisms of justice. Another key factor is the undermining of judicial independence. When courts become politically compromised, they no longer act as a bulwark against authoritarian tendencies. Ultimately, what is legal takes precedence over what is right.

As personal virtue and community responsibility decline, society becomes governed not by shared moral values but by an expanding web of regulations. This growing dependence on the state has not only eroded individual responsibility but also disempowered families and communities, fostering a culture of complacency.

A major sign of democratic collapse is the concentration of power in the hands of a few individuals or a centralized authority. When power becomes too concentrated, it becomes easier for authoritarian figures or political parties to control all branches of government, circumventing checks and balances.

This is often accomplished through election manipulation, control of the media, and the dismantling of opposition parties or civil society organizations that could challenge the status quo. Emergency powers, often granted in times of crisis, are used to justify the centralization of authority and the suspension of civil liberties. Over time, these powers can become permanent, especially when they are justified as necessary for maintaining national security or public order.

Slowly and almost imperceptibly, as democratic institutions collapse or are undermined, there is a growing risk of transitioning from democracy to tyranny. Totalitarian regimes are characterized by absolute state control, the suppression of individual freedoms, and the repression of dissent.

Tyranny often emerges as a reaction to a crisis or societal failure, as leaders promise to restore stability through strong, centralized power. Totalitarian regimes thrive on the elimination of opposition and the centralization of authority. The methods used to suppress dissent can include censorship, imprisonment, forced labour, and even violence.

As totalitarian leaders seek to solidify their control, they often create an atmosphere of fear and surveillance, where dissent is not tolerated, and opposition is crushed. The media is censored or state-controlled, and any alternative viewpoints are branded as subversive or dangerous.

….

By tracing these patterns—economic instability, moral decay, loss of societal cohesion, and the rise of centralized power—we see that they are not separate crises but deeply interconnected. The loss of personal responsibility fuels state expansion, which in turn weakens democratic institutions, making authoritarianism more likely.

As societies grow increasingly dependent on the state and financial speculation, a culture of dependency emerges. Rather than fostering individual initiative and entrepreneurship, modern systems have encouraged reliance on state subsidies, entitlements, and public assistance.

This dependence is not just economic but also cultural—many people no longer see themselves as the architects of their own future but as recipients of the state’s benevolence. This contributes to a sense of learned helplessness and a lack of agency within large segments of society.

At the same time, the rise of a dependent class feeds into class envy, which deepens the divide between the wealthy elites, who have access to resources and power, and the working or underprivileged classes.

Furthermore, the entrepreneurial spirit, which historically drove economic growth and innovation, is stifled by the absence of valid economic calculation. Entrepreneurs are increasingly discouraged from making long-term investments or taking time risks, knowing that state intervention, regulation, lack of legal certainty, and financial speculation may distort markets. The innovative energy of a society is sapped.

….

For individuals and businesses, inflation means that their savings lose value, and it becomes increasingly difficult to make long-term economic plans. The instability caused by inflation leads to uncertainty and a loss of trust in the currency itself. People begin to view money not as a reliable store of value but as a transitory commodity that can be manipulated by those in power.

Furthermore, the more the state intervenes in the economy through welfare programs, the more it needs to print money to sustain those interventions. This creates a vicious cycle: As inflation rises and the value of money declines, the need for more money to fund welfare increases, leading to further inflation and greater dependency. 

This dynamic erodes individual responsibility and leads to a society where personal initiative is supplanted by state control.

The loss of faith in money leads to a breakdown in social cohesion. Money is no longer a trusted medium of exchange or a tool for cooperation; it becomes a source of conflict, with people scrambling to protect their wealth from devaluation.

Speculative behaviour rises as individuals seek alternative stores of value, leading to asset bubbles and further economic instability. Speculation becomes the dominant form of wealth accumulation, rather than productive investment in the real economy

AI progress marches on, here are some useful videos and links:

0:00 Introduction & Going Live 0:43 Meet Palmer: From Baseball to Blockchain to AI 2:24 Overview: AI Development Workflows 3:57 The New Project Setup Document 5:02 Document #1: Project Overview 5:22 Document #2: User Flow 5:45 Document #3: Tech Stack (Most Important) 6:03 Document #4: UI and Theme Rules 6:22 Document #5: Project Rules 7:17 Document #6: Phase Documentation 8:05 Docs Folder Structure 12:36 Cursor Rules vs Cursor.rules File 14:27 User Rules Configuration 18:02 Q&A: Boilerplate vs Starting Fresh 18:50 Custom Modes in Cursor 23:35 Deep Dive: New Project Setup Steps 27:17 AI-First Codebase Design 29:37 Iterative Development with Checklists 35:08 Planning Outside the IDE 38:00 Building with Documents 39:50 Context Management 43:01 Rule Inclusion Hierarchy 45:15 Cursor Notepads Feature 49:30 Chat Management 52:02 Debugging Strategies 54:46 Learning from Errors 57:23 Checkpoint Management 58:51 Closing & Resources

Slides: https://docs.google.com/presentation/d/1kCNuSck8sRpeyaPg1ElgRsMXvweU9XfL1SjO1xUi9DQ/edit?usp=sharing

https://x.com/wadefoster/status/1930680089651425452  “Zapier is measuring PMs on AI fluencyYou only get Capable if you have good prompts & use AI for PRDs and research synth. This is the baseline

AI notes:

Based on the detailed text content about Shopify’s AI-first approach, here is a comprehensive list of practices you can apply at your company to leverage AI effectively and transform your engineering and broader teams:

1. Early and Broad AI Tool Adoption:
   Shopify was an early adopter of AI coding tools like GitHub Copilot (even before commercial availability), and later introduced tools like Cursor and Cloud Code. They experiment with multiple AI tools, evaluating what works before scaling usage.
2. Cross-Functional AI Usage Beyond Engineering:
   AI tools are not limited to engineers; teams in finance, sales, support, and other functions use AI (e.g., Cursor) to build personal productivity tools like MCP servers connecting to Salesforce, Gmail, Slack, empowering non-technical staff to “vibe code” solutions independently.
3. Pairing and Collaborating with AI Labs:
   Shopify’s leadership, including the head of engineering, actively pairs with engineers from AI labs (Anthropic, OpenAI) to understand emerging tools and incorporate best practices, fostering two-way learning and innovation.
4. Building Internal AI Infrastructure (LLM Proxy & MCP Servers):
   They built an internal Large Language Model (LLM) proxy to securely access AI APIs without leaking sensitive company or customer data and manage token usage with visibility and cost controls. They also deploy many MCP (Modular Control Protocol) servers to expose internal and external data sources for AI consumption, enabling easy integration.
5. No Cost Limits on AI Usage:
   Shopify imposes no cost constraints on AI token consumption for engineers or teams, viewing AI tools as productivity investments rather than expenses to be cut. They celebrate high AI usage as a sign of impactful work.
6. Promoting AI Reflexive Culture Through Hiring:
   Shopify massively expanded its intern program (up to 1,000 interns/year) to onboard AI-native talent who naturally integrate AI tools into workflows and help shift company culture toward AI-first thinking.
7. Encouraging Hands-On Leadership and Role Modeling:
   Leaders actively use AI tools themselves and share their workflows and learnings openly, demonstrating the value and encouraging adoption. For example, engineering directors and above are required to do coding interviews, often using AI tools, to emphasize staying close to technology.
8. Allowing Non-Engineers to Build with AI (“Vibe Coding”):
   Non-technical employees are empowered to build small tools using AI without formal engineering intervention, fostering innovation and reducing bottlenecks.
9. Using AI to Reduce Toil and Address Tech Debt:
   Engineers leverage AI to tackle refactoring, tech debt, and infrastructure improvements—tasks traditionally difficult to prioritize due to feature pressures—amplifying engineers’ productivity and system quality.
10. Establishing AI-Enhanced Project Management Tools:
    Shopify built its own project management system (GSD – Get Stuff Done) that integrates AI to automatically draft weekly updates using data from pull requests and Slack conversations, reducing manual reporting toil while maintaining accountability.
11. Setting Clear AI Usage Expectations:
    A company-wide memo established that while AI use is not mandatory, the expectation is that employees should use AI tools to achieve productivity benchmarks, similar to how tools like Excel are assumed available.
12. Transparency and Measurement of AI Impact:
    Usage metrics like token consumption, exception counts, and engineering focus time are tracked and surfaced to managers to monitor AI adoption and team health.
13. Adopting a “Try and Learn” Mindset with AI Tools:
    Shopify encourages experimentation—trying new AI tools or workflows, even if initial results are imperfect or wasteful, with a willingness to discard failed attempts but learn fast.
14. Investing in AI Tooling and Infrastructure Continuously:
    The AI platform and tooling are evolving weekly, with dedicated teams managing AI infrastructure like the LLM proxy and MCP servers, ensuring security, scalability, and ease of use.
15. Combining AI with Human Oversight:
    Despite AI-generated code or content, human review and understanding remain essential—engineers must understand and verify AI outputs before submitting or deploying.
16. Encouraging AI Literacy at All Levels:
    From interns to senior engineers and managers, everyone is encouraged to develop AI skills, with internal prompt libraries and hackathons to share knowledge and accelerate adoption.
17. Hiring and Promoting Leaders Who Embrace Coding and AI:
    Engineering leaders are expected to stay technically engaged and fluent in coding, using AI tools effectively, rather than moving away from hands-on work.
18. Using Specialized AI Tools for Different Tasks:
    Shopify recognizes different AI tools have different strengths (e.g., Gum Loop for web scraping, Cloud Code for coding) and encourages using the right tool for the right job rather than one-size-fits-all.

Based on the provided text, here is a detailed list of all the AI tools, frameworks, platforms, and related technologies mentioned that Shopify uses or has experimented with:

1. GitHub Copilot – Early AI coding assistant integrated into VS Code, used extensively by Shopify engineers since 2021.
2. Cursor – An AI tool/platform used internally by Shopify, growing beyond engineering into teams like finance, sales, and support for building personal productivity solutions. Used alongside VS Code.
3. Claude Code – An agentic AI workflow tool that Shopify adopted and explored deeply, including pairing with engineers from Anthropic to understand internal uses.
4. Anthropic’s Models and Tools – Includes Anthropic’s AI models and Claude code usage; Shopify actively pairs with Anthropic engineers.
5. OpenAI Codex – Referenced as an AI code generation and refactoring tool used for tech debt reduction and infrastructure improvements.
6. Gum Loop – A browser-based automation platform used for tasks like web scraping and monitoring website content, distinct from code-centric tools.
7. LLM Proxy – Shopify’s internally built secure proxy to access various language model APIs (e.g., ChatGPT, Gemini, Claude) while protecting sensitive data.
8. Libra Chat – An open-source native chat product that Shopify contributes to and builds upon for internal AI-powered chat interfaces.
9. MCP Servers (Modular Control Protocol) – A protocol and framework Shopify uses to expose internal and external data/services for AI integration; Shopify runs about two dozen MCP servers (e.g., for Salesforce, Figma).
10. SonarQube (Sonar) – Industry-standard code quality and security analysis tool, including advanced security features, used at Shopify.
11. Work OS / OKIT – Work OS is mentioned as a SaaS app platform that includes OKIT, a user management solution, which Shopify and other startups use.
12. Static (Statsig) – Unified platform for feature flags, analytics, experiments, and more; mentioned as a tool to ship and measure product features.
13. VS Code – The primary integrated development environment (IDE) at Shopify, integrated with AI tools like GitHub Copilot.
14. Devon – An AI tool Shopify experimented with for coding workflows, though results were mixed.
15. Gemini – An AI model/provider mentioned alongside Anthropic, OpenAI, and Claude as part of the AI ecosystem Shopify engages with.
16. Claude – AI model/platform from Anthropic, used and integrated via Shopify’s LLM proxy.
17. Perplexity – Referenced as an internal “internal perplexity” style interface via Libra chat, providing advanced search and knowledge access.
18. Prompt Libraries – Shopify maintains internal prompt libraries for sharing and reusing AI prompts across teams.
19. GSD (Get Stuff Done) – Shopify’s internally built project and product management tool that integrates AI to auto-generate weekly updates using PRs and Slack conversations.
20. Slack – Used as a communication and collaboration channel; AI tools integrate with Slack for context and updates.
21. Jira and Linear – Mentioned as common project management tools Shopify consciously avoids in favor of their own GSD tool.
22. ChatGPT – Widely referenced AI chatbot platform from OpenAI, used in various capacities and integrated securely through LLM proxy.
23. Operator – Mentioned in the context of recording browser sessions to analyze commerce flows, used in AI experiments internally.

Best Practices

1. Experimentation and Iteration
   Companies and engineers should actively experiment with AI tools, sharing tips and tricks internally to discover what works best, as seen in startups like incident.io where teams discuss and refine AI usage collaboratively.
2. Integrating AI with Existing Workflows
   Successful adoption involves embedding AI tools directly into developers’ workflows, such as integrating AI into IDEs, terminals, or code review systems, ensuring seamless use rather than forcing separate tools.
3. Use Well-Defined Tasks for AI Agents
   AI agents perform better with well-scoped, well-defined tickets or tasks, allowing them to generate first passes that developers can then refine, improving efficiency and reducing errors.
4. Automate Repetitive and Low-Value Tasks
   Automation of ticketing, code generation, PR writing, and documentation helps save time and reduces cognitive load, increasing productivity as reported by engineers at Amazon and others.
5. Adopt API-First and Modular Architectures
   Following Amazon’s example, exposing all functionality via APIs enables easier integration with AI agents and automation tools, facilitating smoother AI adoption and extensibility.
6. Maintain Cautious and Incremental AI Deployment
   Big tech like Google emphasizes a slow, careful rollout to build trust and reduce mistrust among engineers, ensuring AI tools are reliable and useful before wide-scale adoption.
7. Encourage Cross-Team Sharing and Learning
   Internal communities sharing AI experiences and strategies foster faster learning and better use of AI tools, promoting cultural acceptance and knowledge dissemination.
8. Balance AI Use with Human Review
   Due to hallucinations and bugs in AI-generated code, continuous human oversight remains critical; engineers still review and correct AI output rather than fully relying on it.
9. Focus on Practical ROI and Time Savings
   Measuring realistic productivity gains (e.g., 3-5 hours a week saved) helps assess the true impact of AI tools rather than inflated expectations, guiding better investment and usage.
10. Treat AI as a New Abstraction Layer
    Recognize that LLMs and AI agents represent a lateral shift across the software stack, enabling new ways to interact with code, data, and tools at all abstraction levels.

Tools

1. Claude Code (by Anthropic)
   A command-line interface tool that sees heavy usage within AI dev tool startups, supporting terminal-based coding and allowing high portions of code to be AI-generated.
2. MCP (Model Context Protocol)
   An open protocol allowing AI agents and IDEs to connect with databases, GitHub, Google Drive, and other services, enabling chat-based querying and automation of workflows.
3. Windsurf (AI IDE Editor)
   An AI-powered editor where up to 95% of the code is generated or assisted by AI tools, supporting coding through agent interaction and passive tabbing.
4. Cursor
   An AI coding assistant with approximately 40-50% code generation success, representing a more balanced but honest approach to AI coding productivity.
5. Google’s Internal Tools
   • Cider: Google’s cloud-integrated development environment built on a VS Code fork.
   • Critique: AI-powered code review feedback tool.
   • Code Search with LLM support: Advanced search tool enhanced with AI for querying codebases.
   • Notebook LM & LM Prompt Playground: AI chat-based documentation and prompt experimentation tools.
   • MoMA Search Engine: Knowledge base powered by LLMs for internal information retrieval.
6. Amazon Q Developer Pro
   An internal AI tool specialized for AWS-related development, praised for its contextual awareness and coding assistance within the AWS ecosystem.
7. AI Agents for Automation
   Tools that automate ticketing, emails, internal workflows, and code deployment pipelines, heavily used within Amazon and other organizations for increasing efficiency.

https://www.amplifypartners.com/blog-posts/the-agent-first-developer-toolchain-how-ai-will-radically-transform-the-sdlc

• Features for an agent-first IDE
  • An intent specification layer. Agents don’t just need prompts—they need structured, machine-readable intent representations. The IDE must support DSLs for describing desired outcomes, tools to define constraints like budget limits and performance SLAs, and the ability to compare current system state to desired goals, and let agents plan accordingly.
  • Agent-framework integrations. Instead of plugins for tools like linting and debugging, future IDEs will integrate with agent frameworks—systems like Cody, Devin, Claude Code, or custom LLM pipelines. These integrations will support UIs for configuring agent roles and permissions, interfaces for visualizing and steering how different agents (e.g. frontend, backend, infra) are collaborating, and persistent agent memory that allows long-term project context across sessions.
  • Agent monitoring and telemetry. Agents need constant monitoring to ensure correctness, safety, and alignment. We will need live dashboards to visualize agent tasks and status, trace-based debugging to step through an agent’s reasoning chain and code generation path, and alerts on unexpected behaviors like unsafe code or violating performance bounds.
  • Explainability and decision logging. In a world of autonomous agents, trust depends on transparency and fine-grained auditability; more on this below. Agents will need to output action logs for each input, output, and rationale. We will want interactive code diff explanations (“why did this code change?”), plus some basic system summaries: the ability for agents to overview architecture, data flows, etc.
  • Automated testing & verification. This is the subject for an entire other blog post, but the IDE of the future needs to be able to verify agents. This verification needs to cover code tests (unit, integration, etc.), and support simulated sandbox environments to validate code behavior before merging. And finally, IDEs will need to support formal verification methods like TLA+ and ROCQ.
  • Live collaboration and feedback loops. For humans and agents to work well together, we’ll need real-time synchronization via real-time agent-human chat (like Copilot chat, but system wide), editable simulations that let you preview a feature and leave feedback (like in Google Docs or Figma), and live agent tuning via adjusting hyperparameters, toolsets, or constraints.
  • Knowledge graph and memory integration. Agents will need structured knowledge about the system, business logic, and users. I expect to see things like a semantic project graph, where the codebase (functions, models, services) is represented as an interconnected knowledge graph. We will also need embedded documentation plus persistent long-term memory that allows agents to remember past decisions, tradeoffs, and user preferences.
  • A pluggable execution environment. The IDE of the future needs a sandbox where agents can safely run and test code. Examples include containerized runtime environments, policy enforcement engines to restrict what agents can deploy or access, and a tooling-calling interface layer that lets agents use CLI tools, APIs, and libraries with traceable logs.
• Think of this new IDE as a mission control center for intelligent systems; the developer becomes a strategist, curator, and verifier rather than a code-writer.

Key Insights

• 🤝 Human-AI Collaboration Redefines Work: The future of work will revolve around humans deploying, managing, and reviewing AI agents rather than executing every task manually. This represents a fundamental shift in how knowledge work is performed, requiring new skills focused on orchestration and quality assurance rather than rote execution. The probabilistic nature of AI means human oversight remains crucial to avoid critical errors, emphasizing a hybrid human-AI workflow.
• ⚙️ AI as Capability Expansion, Not Just Cost-Cutting: Box’s approach contrasts with companies focused purely on headcount reduction. Instead, AI is used to amplify what employees can accomplish, enabling faster delivery of products, more tailored marketing campaigns, and enhanced customer support. This mindset creates a virtuous cycle where efficiency gains lead to reinvestment and growth, rather than mere downsizing.
• 📈 Jevons Paradox Applied to Labor: The speaker applies the Jevons paradox—where efficiency gains can increase overall consumption—to labor. By making tasks more efficient, AI can increase demand for services that were previously too costly or slow, such as healthcare consultations or legal reviews, leading to growth in employment in those sectors rather than contraction.
• 🏢 Enterprise AI Adoption is a Marathon, Not a Sprint: Despite rapid AI innovation, real-world organizational adoption is slow and complex due to change management challenges. Companies may take years to fully integrate AI agents into workflows, similar to how cloud adoption took decades to mature. This highlights the importance of patience and strategic planning in AI deployments.
• 🌍 New Markets and Industries Will Be Digitized: Many professional services industries have historically been under-digitized due to the nature of their work. AI agents will catalyze the digitization of sectors like legal services, healthcare, education, and drug discovery, creating new markets that were previously inaccessible to software solutions. This opens vast opportunities for startups and incumbents alike.
• 🔄 Ecosystem Diversity and Interoperability are Key: The future AI landscape will not be dominated by a single player or approach. Instead, a diverse ecosystem of AI model providers, SaaS vendors, and agent frameworks will coexist. Interoperability standards such as ADA and MCP will be critical to enable agents to communicate and collaborate across platforms, enhancing flexibility and innovation.
• 🛡️ Ethical and Governance Challenges Are Front and Center: Responsibility for AI agent outputs remains with humans for the foreseeable future. As AI systems become more autonomous, new frameworks for liability and governance will emerge, possibly holding AI providers accountable. Additionally, concerns about AI-generated misinformation and the amplification of echo chambers require proactive strategies to maintain trust and information quality.
• 💬 The Future of Software UX and Agent Integration: While some speculate that AI agents might replace traditional user interfaces, the speaker believes that most users still prefer deterministic, dashboard-style interfaces for routine interactions. Agents will operate in the background or within SaaS platforms, handling complex tasks and workflows, while users engage with familiar interfaces enhanced by AI.
• 🧠 AI Memory and Personalization Pose Complex Trade-offs: Integrating personal AI memories with corporate AI systems raises significant privacy, governance, and IP challenges. Maintaining separation (“church and state”) between personal and work agents may limit the full potential of personalized AI but is necessary to manage risks. The industry may benefit from open standards around agent memory portability to facilitate interoperability.
• ✍️ Human Creativity and Judgment Remain Irreplaceable: Despite AI’s ability to generate text, code, and other outputs, humans still need to critically assess, edit, and contextualize AI-generated content. The speaker continues to write and think independently to maintain understanding in a rapidly evolving landscape, underscoring that AI is a tool to augment, not replace, human insight.
• 🚀 Startups Have a Unique Window of Opportunity: Unlike horizontal AI assistants dominated by large players, many newly digitized industries have no entrenched incumbents, giving startups an opportunity to capture market share with innovative AI-driven solutions. This dynamic mirrors past waves of disruption seen with companies like Uber and Airbnb and suggests a fertile environment for AI entrepreneurship.
• 🔍 AI Agents Will Drive Massive API Growth and Browser Use: The rise of AI agents will fuel a surge in API creation as software vendors expose more programmable interfaces to enable agent integration. However, some tasks will always require browser-like interactions where APIs cannot fully capture the needed complexity, ensuring agent-based browser automation remains relevant.
• ⚠️ Managing AI-Generated Content Quality is a Critical Challenge: The proliferation of AI-generated content risks overwhelming information ecosystems with low-quality or misleading material (“AI slop”). Verifying human-generated content and mitigating echo chamber effects will become essential, especially as agents mediate more of our interactions with digital information.
• 🕰️ Change Management and Adoption Pace Are the True Bottlenecks: The speaker emphasizes that while AI technology evolves at a breakneck speed, the real constraint is how quickly individuals and organizations can adapt workflows and mental models to incorporate AI effectively. This process involves iterative experimentation, learning, and cultural shifts that take significant time.

12 years ago I posted speeches by Pierre Poilievre and Mark Carney, the two front runners to the current federal election. Both speeches focused on getting Canada’s house in order, so its interesting to reflect on them now.

I compared Mark Carney‘s speech ‘Growth in the age of deleveraging‘ (2011) with the Liberal party platform, using AI (o3), with some edits and added charts below:

• Bottom Line: Mark Carney’s 2011 speech called for deleveraging before the next shock hit. Instead, every sector has levered further in the intervening 14 years. That makes the Liberal platform’s reliance on larger deficits a higher-stakes wager than it would have been in 2011: the payoff must be faster productivity growth, or Canada enters the next downturn with far thinner buffers than the Bank of Canada once recommended.
• The platform emphasizes sovereignty, housing affordability and the energy transition — objectives that are hard to achieve quickly without public balance-sheet support.
• Reframing the risk. Carney’s 2011 concern was too much household leverage financed at teaser rates. The platform recasts the main risk as external shocks (U.S. tariffs, geopolitical fragmentation) and argues that public investment can crowd in private capital and lift the growth denominator faster than the debt numerator.

Snapshot of the two documents

Where the two visions align

1. Productivity & investment are Canada’s long-run growth lever.
   – The 2011 speech urges business to launch a “virtuous circle of increased investment and increased productivity.” Bank of Canada
   – The 2025 platform likewise chases capital formation—just via heavy federal spending and targeted sector plans. Liberal Party of Canada
2. Trade diversification beyond the United States.
   – Carney (2011) tells firms to “access fast-growing emerging markets.” Bank of Canada
   – The platform proposes east-west trade corridors, Arctic ports and high-speed rail to cut internal costs and open global routes. Liberal Party of Canada
3. Acknowledgement that demographics and productivity trends have turned less favourable and need policy attention. Bank of Canada Liberal Party of Canada

Where they diverge

A concise synthesis

Back in 2011, Governor Carney argued that the post-crisis world demanded prudence and private-sector-led rebalancing: pay down debt, lift business investment, find new export markets, keep the state’s balance-sheet dry.

In 2025, the Liberal platform he now fronts adopts a nearly mirror-image response to a different shock: large-scale public investment and a more interventionist federal hand to defend sovereignty, accelerate clean growth, and cushion households.

Both documents share a core belief that productivity and diversified trade are Canada’s path to prosperity. What has flipped is the chosen vehicle—fiscal activism vs. fiscal restraint—and the perceived threat—external tariffs today vs. global leverage yesterday.

Debt Metric Recap:

While friends were constantly debating the Israel / Palestine conflict, which was constantly receiving much attention, in Nov 2023 I thought I would remind them of the big picture in the words of Peter Zeihan: The world is heading towards multipolarity

“Listened to a couple zeihans, took some notes, here it goes:

Peter Zeihan – Book: The end of the world is just the beginning – tilt is bearish

CHINA

• Last 10 year period. Nation collapse. Demographics decline , can it be managed/mitigated?
• China has size, us industry move out as China 3-4× faster than went in.
• You can’t climb into someone’s head when you don’t know what they know (Xi gets limited information).
• China: how China disintegrates is order of magnitude more important than Ukraine. Worlds workshop.35 trillion sunk cost industrial base. How that process happens will be crucial.

RUSSIA/Ukraine. 

• Ukraine is order of magnitude more important than Gaza. Gaza is a distraction. If Ukraine falls have to deal with rest of Europe, If Russia falls and we have to deal with the disintegration of Russia. Structure of region and state of NATO alliance at risk.

Multipolarity / Security Guarantees

• If countries think that us won’t guarantee their security, they might create/source nukes. Sweden, Poland, South Korea, etc.
• Coming out of ww2 they understood the need for intervention/guarantee security/police sea lanes/pax americana,, not anymore 
• Missle defense crucial 

Israel/Palestine:

• US has a social contract with its citizens, whoever you are, wherever you are we will come get you.The U.S. has one thing on its mind — get our people out — everything is second to that. 
• This region lacks a broad geopolitical significance that may otherwise entice external involvement, which means this conflict will remain an isolated issue for Hamas, Gaza and Israel
• Interrupt Normalization of relations between 2) Suni, Saudi arabia uae and 3.) Israelis jews. Ideas was if you can get these factions on the same side, the rest of the middle east would take care of themselves

Follow up: cross reference with Marko Papic, ‘stratfor mafia’

Marko Papic

• Geopolitics and Trade
• https://www.youtube.com/watch?v=dGmjY4OthxI&list=PLNAjSv6RDKY82-T3_8qSpcyopR-VNU8H5&index=21 Episode 7: A New Global Order: Integrating Geopolitics with Anastasia Titarchuk and Marko Papic
• https://x.com/Geo_papic/status/1740075876472299638

Source: Deaths in state based conflicts by region

Other previous commentary from Peter Zeihan

Disunited Nations by Peter Zeihan

 “The world has gotten so used to the “normal” of an American-dominated Order that we all have forgotten the historical norm: several smaller, competing powers and economic systems throughout Europe and Asia. The period of American hyper-involvement in global affairs is ending. The impacts on global energy and agricultural markets, finance and technology will be transformative, but the heirs to the dying Order are not who you think. Russia, India, China and Brazil will not be the superpowers and wunderkind of the future. Rather, names familiar and new will be taking charge of the emerging global Disorder. “

“Most countries and companies are not prepared for the world Peter Zeihan says we’re already living in. For decades, America’s allies have depended on its might for their economic and physical security. But as a new age of American isolationism dawns, the results will surprise everyone. “

https://twitter.com/PeterZeihan/status/1497977611464572930?s=20

“Russia: Outlook: Russia is an aging, insecure, former power determined to make a last stand before it is incapable of doing so. American disengagement from the global scene couldn’t have come at a better time, but the reactivation of Russia’s traditional local foes couldn’t have come at a worse one.” (Demographics and Geography are against them)

Minerals used in Green Technologies (Transport and Power Generation). Replacing 1 complicated supply chain with 13 others…

The alpha, your job as an AI practitioner, is to ask “What will everyone be doing in 1yr that I can do now?”

•Force yourself to use AI, think of how you can automate something when you’re doing it

•Don’t go for your old routine. Use AI and Low/NoCode. It won’t be convenient at first, but it’ll force you to spend most of your time doing what actually matters in the future?

•What’s your AI aha moment? ChatLLM? NotebookLM? Illuminate? Gemini 1.5 + Deep Research? Cursor? Bolt? ChatGPT? Etc…•

“If you’re an …. your goal should be to move to the forefront of what is made possible by AI and just ride that wave. A time like now when something is young and is blowing up but experts are few and best practices are ill-defined are when you can find absurd amounts of alpha. Honestly, the delta between what the current ….. knows and the forefront of AI is smaller than most assume, but I’m shocked at how few …… are working that direction. …. are too burned out from hype cycles that didn’t come to pass, but this time is different.”

“There is a MASSIVE CHASM between ai native and non ai native people that will be filled with LEARNING more than software.”

“”Vibe knowledge work” could mean a way of working where you rely on intuition, creativity, and AI assistance to manage, process, or generate knowledge, rather than getting bogged down in rigid structures or manual effort. Imagine using natural language to direct AI tools to research, summarize, or connect ideas for you—focusing on the big picture or the “vibe” of what you’re trying to achieve, like understanding a topic deeply or crafting a compelling argument. The AI would handle the tedious details (like sifting through data or formatting reports), while you steer it with high-level intent, maybe even through casual prompts or voice commands.”

@rowancheung

•Sam Altman on what you need to do to survive in the age of artificial intelligence.

•”You are about to enter the greatest golden age of human possibility…”

•To thrive in that world, the skills that matter most are:

•– Deep familiarity with the tools

•– Staying abreast of changes

•– Developing a great intuition for AI tools, where things are going, and how to make use of it

•– Resilience and the ability to learn things fast and evolve yourself with technology

•I know most of this stuff is a pretty big no-brainer for anyone paying attention, but here’s the takeaway:

•AI upskilling and keeping up with AI are possibly the most important skills in the world right now.

•And the most fascinating part is that new AI developments and tools come out so fast that everyone is constantly learning.

@Francis_YAO_

•I’ve been asked by few first year PhD about how to start LLM research on X, say long context modeling. My number one suggestion — though it seems a bit of unconventional — is *not* to read any papers related to long-context, but to talk to the model – Talk to the model about a text book, course slides, financial reports, novels, nonfictions, any long document you could find – Talk to the model for two whole weeks, from the morning first thing after opening up the laptop, to the evening last thing before going to the bed. – Ask every single question you could imagine, what is PCA? How does it compare to SVD? Which part of the book describes the two? What the book says exactly? – Talk to all the models you could access, GPT, Gemini, Claude, Llama … – Keep talking to the model for two whole weeks, no research, no paper, no arxiv, just talk to the model. – During the above process, continuously observe how the model behave, discover their problems, and think about why models could behave that way I found people who have gone through the above process have a fundamentally different level of understanding than people who just read papers

•If you want to do research on AI, or figure out how it can be used in your organization, the first step is talk to the models a lot. Use it for everything you do (within legal & ethical bounds). You don’t know what it does until you use it.

@johnrushx

•The Future of <Software Developer> profession:

•> software developer jobs will mainly become obsolete

•> product builders will replace them & be huge

•> code itself eventually will be abstracted far away by AI builders + component/block libraries + nocode/lowcode

•> the number of solo product builders will grow from less than a million to 10s of millions.

•> 1% of the best devs will be building these platforms & legacy/corporate/critical software

•> 99% will be forced to become “product builders”

•> most will ignore this until it’s too late and they’re out of their jobs with no skills to get a new one on same pay grade

•>> The best way to prepare for the change is to become a part-time indie hacker. Don’t quit your job; build and ship pet projects in the evenings and weekends.

•When you do so, don’t go for your old routine.

•Use AI and NoCode. It won’t be convenient at first, but it’ll force you to spend most of your time doing what actually matters in the future:

•> UX

•> ideations and validation

•> content creation

•> learning to win attention

•> translating pain points into products

•> training your eye on patterns

•> mastering productivity

•> social media

https://www.oneusefulthing.org/p/innovation-through-prompting

https://www.oneusefulthing.org/p/detecting-the-secret-cyborgs?r=i5f7

Four Questions to Ask About Your Organization.

So how can leaders start to think about the rapidly advancing nature of AI? The first thing they should do is use it. No amount of reading and research can substitute for spending 10 hours or so with a frontier model, learning what it can do. After getting familiar, companies should think about the following four questions:

1.What useful thing you do is no longer valuable? AI doesn’t do everything well, but it does some things very well. For many organizations, AI is fully capable of automating a task that used to be an important part of your organizational identity or strategy. AI comes up with more creative ideas than most people, so your company’s special brainstorming techniques may no longer be a big benefit. AI can provide great user journeys and personas, so your old product management approach is no longer a differentiator. Getting a sense of what AI can do now, and where it is heading, will allow you to have a realistic view of what might soon be delegated to an LLM.

2.What impossible thing can you do now? The flip side of the first question is that you now can do things that were impossible before. What does having an infinite number of interns for every employee get you? How does giving everyone a data analyst, marketer, and advisor change what is possible? You can look at some of the GPTs my students created as inspiration.

3.What can you move to a wider market or democratize? Prior to AI, companies were often advised to put their effort into servicing their most profitable customers, but AI has greatly changed the equation. Services and approaches that were once expensive to customize have become cheap. Prior to AI, strategy consulting firms would only work for giant clients for large fees, but now they may be able to offer effective advising to a much wider range of businesses at lower costs. Custom tutoring and mentoring, once available only to the rich, may be widely democratized.

4.What can you move upmarket or personalize? At the same time, your organization’s capabilities have increased. If you were once a small marketing firm, you can use AI to punch above your weight and offer services to elite clients that were once only available from much larger firms. With giant context windows and fast answers, every customer may be able to have a personal AI agent who knows their preferences and previous interactions with the company and communicates with them according to their preferences. Figure out the most exciting thing you can do, and see if you can make it happen.

Misguided companies will see any increase in performance from AI as an excuse to lay off staff, keeping their output the same. More forward-thinking firms will take advantage of these new capabilities to both improve the lives of their employees and expand their own capabilities. This is an area where leaders have agency over the future of AI and work. A lot depends on getting it right, and fast, because it is possible we are just getting started.

https://X.com/clairevo/status/1814747787856388435

https://www.mindstone.com/programs/ai-competency

https://www.aitra.com/contact-us

@levie
AI is going to cause us to move to higher levels of abstraction of how we work. Each level of abstraction provides more leverage than the prior level, so each bit of input leads to vastly higher output.

This has happened all throughout history when there’s major technological progress, from the Industrial Revolution with mechanical automation and in the Information Age with digital automation. The work that we do today looks far different from 100 or 50 years ago respectively.

The same will be true again with AI. What we perceive is “work” today will continue to be redefined. When you can merely think of an idea to prototype and AI can generate the code, the timelines on building software suddenly alter. When you can instantly research a topic and understand it deeply without the hundred hours going down the wrong threads, you’ll move to the next task much quicker.

This will naturally change what we spend our time doing each day in almost every field. Building software will be as much about reviewing code and considering the right prompts as it is writing the code. Delivering a healthcare outcome will mean having access to every bit of research at your fingertips instantly, augmenting anything you already know. Every domain will experience a similar impact.

Skills will matter just as much as ever, but they will look different, just as skills have changed during every other technological revolution. And more people can get started in a field they’re interested in, while the experts in the field can get even more done than they could’ve before.

In just a few years, we will look back on how we used to work and be utterly surprised how long everything took to do. It will seem implausible that you had to literally do everything yourself on a computer, the thing that was invented to help automate work.

How I use LLMs