A Generalist #3 – MCP, Ethics, Shopify, Intel and TSMC
Welcome to the third edition of A Generalist, your monthly digest of technology, product development, cloud computing, marketing, travel, outdoor sports, camping, books, and a touch of politics.
The AI Revolution: Context, Control, and Controversy
Decoding the Model Context Protocol (MCP): The USB-C for AI?
A significant development in the realm of artificial intelligence integration is the emergence and growing adoption of the Model Context Protocol (MCP). This open standard aims to streamline how large language models (LLMs) connect with external data sources and tools, offering a more unified and efficient approach compared to previous methods. Envisioned as a kind of "USB-C for AI," MCP provides a universal protocol that allows AI assistants to securely access information from various repositories, business applications, and development environments.
Anthropic, a leading AI research company, played a key role in open-sourcing MCP, highlighting its potential to standardize the interaction between AI systems and the vast amounts of data they need to function effectively. This standardization simplifies the complex web of integrations that would otherwise be required for each new data source an AI model needs to access. Instead of numerous custom connections, developers can build against a single, consistent protocol. The benefits of MCP are manifold. It offers real-time access to information, eliminating the issue of outdated responses that can plague AI systems relying on pre-indexed data. Furthermore, it enhances security and control over sensitive information, as data is accessed directly without the need for intermediate storage or processing like embeddings. The protocol also promises to lower the computational load associated with data retrieval, leading to greater efficiency. Its inherent flexibility and scalability make it an attractive solution for companies operating across diverse platforms and databases.
The ecosystem surrounding MCP is rapidly expanding, with curated lists of available MCP servers being actively maintained on platforms like GitHub. These lists showcase a wide array of servers designed to connect AI models with various applications and services, including platforms like YouTube and data management tools like Redis. From a developer's perspective, MCP offers a relatively straightforward way to build and utilize these connections, with available SDKs and guides simplifying the process. This ease of use has the potential to empower more developers to integrate sophisticated AI capabilities into their applications, leading to a broader range of innovative AI-powered solutions. The open and collaborative nature of MCP suggests a future where AI interactions are more seamless and accessible, fostering greater innovation across the technological landscape.
Substant AI: Your Personal AI Intelligence Briefing
In an era of information overload, tools that can help individuals efficiently monitor and understand the data relevant to their interests are increasingly valuable. Substant AI emerges as one such platform, leveraging the power of artificial intelligence to provide personalized intelligence briefings. This service is designed to help users stay informed about the companies, stakeholders, and topics that matter most to them by continuously monitoring hundreds of sources.
At its core, Substant AI offers AI-written executive briefs that are tailored to each user's specific interests. These briefs are proactive, ensuring users receive timely updates and never miss critical information. The content is not only personalized but also aims to provide actionable insights, sometimes even suggesting ways to respond to the information received, such as drafting messages.
The Ethics of Idle AI Talent: Google's Controversial Strategy
The intense competition for top talent in the artificial intelligence field has led to some unconventional strategies, as evidenced by reports concerning Google's (DeepMind) approach to retaining its AI expertise. Allegations have surfaced suggesting that Google is paying certain AI staff in the UK to essentially do nothing for periods of up to a year after they leave or indicate their intention to depart the company. The primary motivation behind this strategy appears to be to prevent these highly sought-after individuals from joining rival firms like OpenAI, Meta, or Microsoft during a time when the race for AI dominance is particularly fierce.
This practice reportedly involves the enforcement of non-compete agreements, which are more common and legally binding in the UK where DeepMind is based, coupled with what is known as "garden leave". During this garden leave, employees continue to receive their full salary but are not required to work for Google and, crucially, are prohibited from joining competitors for the duration of the non-compete period, which can extend up to a year. This strategy has drawn considerable criticism within the tech industry, with some former DeepMind leaders and others in the field denouncing it as an "abuse of power". Critics argue that such extended periods of forced inactivity can stifle innovation by keeping talented individuals away from active research and development, and can also negatively impact the career growth of the employees involved. This situation highlights a significant difference in the legal landscape surrounding non-compete agreements between the United States and the United Kingdom, with the US having moved towards banning or restricting most such agreements, particularly in states like California. Google, while acknowledging the reports, maintains that its employment contracts are in line with market standards and that non-compete clauses are used selectively to protect legitimate business interests, especially concerning sensitive projects. The willingness of a major tech company to adopt such measures underscores the extreme value placed on top AI talent and the intense competition to maintain a technological edge in this rapidly evolving field.
Shopify's Reflexive AI Mandate: AI as the New Baseline
In contrast to the more defensive strategy reportedly employed by Google, Shopify has taken a proactive and assertive stance on the integration of artificial intelligence into its operations. A recent memo from Tobi Lütke, the CEO of Shopify, emphasizes that "reflexive AI usage is now a baseline expectation for everyone at Shopify". This is not merely a suggestion or an invitation to experiment, but a fundamental requirement that applies across all roles within the company, including the executive team.
The rationale behind this mandate is rooted in the belief that AI acts as a powerful multiplier, significantly enhancing productivity and enabling employees to achieve results that were previously considered implausible. Lütke draws a parallel to the company's top performers who deliver "10X" results, suggesting that AI has the potential to make the tools themselves "10X" more effective. He has observed instances where the brilliant and reflexive use of AI has led to "100X the work done" on certain tasks. This expectation is tied to Shopify's core values of being a constant learner and thriving on change, with Lütke stating that opting out of learning how to apply AI in one's craft is not a feasible option and will likely lead to stagnation. To ensure this integration, Shopify intends to incorporate AI usage into its performance and peer review questionnaires, indicating that proficiency in using AI will be a factor in evaluating employee performance. While the learning of AI skills is considered self-directed, employees are expected to share their learnings, successes, and even failures with each other. Shopify provides access to a range of cutting-edge AI tools and platforms to facilitate this process. Furthermore, before teams can request additional headcount or resources, they must now demonstrate why their goals cannot be achieved through the effective use of AI, prompting them to consider how autonomous AI agents could contribute to their work. This company-wide mandate signals a fundamental shift in how Shopify operates, with AI becoming deeply ingrained in its culture and workflows, reflecting a strong belief in the transformative potential of AI to redefine entrepreneurship in a world where AI is universally accessible.
Key Insights from the Stanford AI Index Report 2025
The Stanford HAI AI Index Report 2025 provides a comprehensive, data-driven overview of the current state and trends in artificial intelligence, offering valuable insights into its rapid development and increasing impact across various domains. The report highlights the significant improvements in AI performance across demanding benchmarks, with scores on tests like MMMU, GPQA, and SWE-bench showing dramatic increases within the past year. This rapid progress underscores the accelerating capabilities of AI systems.
Furthermore, the report indicates a clear trend of AI becoming increasingly embedded in daily life, moving beyond research labs into practical applications in sectors like healthcare and transportation. The FDA's approval of a significantly higher number of AI-enabled medical devices in 2023 compared to previous years serves as a concrete example of this integration. Similarly, the growing prevalence of self-driving car services points to the tangible impact of AI in everyday mobility. The report also emphasizes the strong business investment in AI, with the private sector in the U.S. reaching record levels of funding in 2024, and a substantial increase in the adoption of AI by businesses across various industries. While AI-related incidents are on the rise, the report notes that the adoption of standardized Responsible AI (RAI) evaluations by major industrial model developers remains limited, though new benchmarks for assessing factuality and safety are emerging. Interestingly, global optimism regarding AI products and services is increasing overall, but with notable regional differences, suggesting that cultural and societal factors play a role in shaping public perception of AI. The report also points to the growing efficiency, affordability, and accessibility of AI, driven by the development of increasingly capable smaller models and declining hardware costs. Finally, the Stanford AI Index highlights the increasing involvement of governments in AI through regulation and investment, as well as the expansion of AI and computer science education in many countries. These key findings collectively paint a picture of a rapidly advancing field that is becoming more integrated into our lives, attracting significant investment, and prompting increasing attention from both businesses and governments worldwide.
Tech Industry Tides: Regulation, Partnerships, and Platforms
Revolut's €3.5 Million Fine: A Regulatory Wake-Up Call
The fintech sector, known for its rapid innovation, is also subject to increasing regulatory scrutiny. A recent example of this is the €3.5 million fine imposed on Revolut by the Central Bank of Lithuania for failures in its money laundering prevention measures. The fine was a result of a scheduled inspection that identified shortcomings in how Revolut monitored business relationships and customer transactions, leading to instances where the bank did not always properly identify suspicious monetary operations.
While the investigation did not uncover any confirmed cases of money laundering, the regulatory body emphasized the need for improvements in Revolut's existing controls. This penalty marks the largest fine levied against the UK-based fintech company to date, underscoring the seriousness with which regulators are addressing anti-money laundering (AML) compliance within the rapidly growing fintech industry. Despite the significant fine, Revolut has stated its commitment to the highest regulatory compliance standards and has cooperated with the Lithuanian central bank to take immediate action to rectify the identified procedural deficiencies. This incident serves as a reminder of the critical importance of robust compliance frameworks for fintech companies as they scale and navigate the complexities of financial regulations across different jurisdictions. The increasing scrutiny from regulatory bodies highlights the need for fintech firms to prioritize and continuously invest in their AML controls to ensure they can effectively detect and prevent illicit financial activities.
Intel and TSMC: A Semiconductor Saga of Partnership and Denial
The semiconductor industry, a cornerstone of modern technology, is often a stage for significant partnerships and strategic maneuvers. Recent reports suggested a potentially groundbreaking collaboration between two of its giants: Intel and Taiwan Semiconductor Manufacturing Co. (TSMC). These reports indicated a preliminary agreement had been reached for a joint venture that would see TSMC operate Intel's chipmaking facilities in the US, with TSMC even taking a 20% stake in this new entity. The US government was reportedly a key driver behind these discussions, aiming to revitalize Intel, which has faced manufacturing setbacks in recent years, and to bolster domestic chip production capabilities considered vital for national security.
However, this narrative took a turn when TSMC publicly denied any involvement in talks about forming a joint venture with Intel. During a presentation of the company's quarterly figures, TSMC Chairman C.C. Wei explicitly stated that TSMC was not engaged in any discussions regarding a joint venture, technology license, or technology transfer with Intel or any other company. This denial came as a setback for Intel, which has been striving to regain its footing in the semiconductor market amidst increasing competition. The discrepancy between the initial reports and TSMC's firm denial underscores the sensitive nature of such high-level discussions in the technology industry and the potential for market speculation to outpace or misrepresent actual developments. Despite the denial, the fact that such a partnership was even rumored highlights the ongoing shifts and potential realignments within the semiconductor landscape, driven by factors such as manufacturing prowess, technological advancements, and geopolitical considerations.
Spotify's Ad Revolution: AI and Automation Take Center Stage
Spotify, a leading audio streaming platform, recently unveiled significant updates to its advertising business, signaling a major push towards leveraging artificial intelligence and automation. At a recent event in New York City, the company announced the launch of the Spotify Ad Exchange (SAX), a new programmatic advertising system. This exchange allows advertisers to reach Spotify's engaged and logged-in user base through real-time auctions, promising greater efficiency and data-driven ad placements. Spotify has partnered with major advertising platforms to power SAX, including Google's Display & Video 360, Magnite, and Yahoo DSP, with more integrations planned.
In addition to this move towards programmatic buying, Spotify is also integrating generative AI directly into its advertising tools. Through Spotify Ads Manager, marketers in the US and Canada can now use AI to create scripts and voiceovers for their audio ads at no additional cost. This initiative aims to streamline the ad creation process, making it easier for advertisers of all sizes to produce high-quality and engaging audio content. These updates reflect Spotify's broader strategy to modernize its ad platform, making it easier for advertisers to buy, create, and measure their campaigns on the platform, with a particular focus on reaching the valuable Gen Z demographic. By embracing AI and automation, Spotify is positioning itself as a key player in the digital advertising landscape, offering innovative tools to help brands connect more effectively with their target audiences.
Redis Embraces AGPLv3: A Win for Open Source?
The licensing of open-source software is a critical aspect of its development and adoption. Redis, a popular in-memory data structure store, recently announced a significant change in its licensing strategy, adopting the AGPLv3 (Affero General Public License version 3) for Redis 8. This move marks a renewed commitment to the principles of open source, particularly after a previous shift to the SSPL (Server Side Public License), which was not universally recognized as a truly open-source license by the Open Source Initiative.
A primary driver behind the adoption of the AGPLv3 license is to address the ongoing challenge of cloud providers profiting from open-source projects without making proportional contributions back to the community. The AGPLv3 license is designed to ensure that if a modified version of the software is used to offer a service over a network, the source code of that service must also be made available under the same license. This aims to foster a more equitable ecosystem where cloud providers who benefit from open-source projects also contribute to their sustainability. This licensing change also serves to unify the developer experience by integrating the advanced features of Redis Stack, such as JSON, Time Series, and probabilistic data types, directly into the core Redis 8 under the AGPLv3 license. This eliminates the need for separate distributions and provides developers with a more comprehensive set of tools within the open-source version. By aligning with a widely recognized open-source license, Redis aims to strengthen its relationship with the developer community and encourage further contributions to the project, ensuring its continued innovation and growth. This decision reflects a broader trend within the open-source world to find sustainable models that balance the collaborative spirit of open source with the economic realities of cloud-based service offerings.
Building and Selling in the Modern Landscape
Zach Perret's Wisdom: Prioritizing Product and Distribution at Plaid
In the dynamic world of technology startups, the ability to not only create compelling products but also to effectively get them into the hands of users is paramount. Zach Perret, co-founder and CEO of Plaid, a company that builds the infrastructure for consumers to digitally interact with their bank accounts, shared his insights on this critical balance in an article titled "Build Products, Sell Products". Perret emphasizes that Plaid operates with two core priorities: first and foremost, to build products that customers actively use, and secondly, to ensure the effective distribution of these products to the market.
According to Perret, the focus on building products that customers use involves a deep understanding of their needs, active listening to their feedback, and an iterative approach to development based on actual customer usage. Simply building what customers say they want is not sufficient; the ultimate measure of success is whether they actually integrate the product into their workflows. The second priority, effective distribution, acknowledges that in the business world, the best product does not always win based on merit alone. Factors such as existing relationships, pre-established agreements, brand reputation, ease of integration, pricing strategies, and bundling all play significant roles in purchase decisions, particularly in the B2B sector. Perret stresses that at Plaid, everyone within the company, across all functions and teams, should be aligned in service of these two core objectives: building great products and distributing them effectively. This philosophy underscores the interconnectedness of product development and go-to-market strategies, highlighting that success in the tech industry requires excellence in both creation and delivery.
Looking Ahead
The technological landscape of May 2025 reveals a confluence of powerful trends. The integration of AI is no longer a futuristic concept but a rapidly evolving reality, as seen in the standardization efforts around MCP, the personalized intelligence offered by platforms like Substant AI, and the varying strategies companies are employing to secure and leverage AI talent. Regulatory bodies are also playing an increasingly active role in shaping the industry, as highlighted by the fine against Revolut. Meanwhile, strategic partnerships and platform innovations, such as Spotify's advancements in ad technology and Redis's licensing shift, demonstrate the continuous efforts to adapt and thrive in a competitive market. The wisdom shared by industry leaders like Zach Perret underscores the enduring importance of focusing on both product excellence and effective distribution. As we navigate these currents of change, a generalist perspective becomes invaluable, allowing us to connect seemingly disparate events and gain a more holistic understanding of the forces shaping our future. The rapid pace of innovation suggests that the coming months will undoubtedly bring further developments and challenges, making continuous learning and adaptation essential for staying informed and engaged.
Book, Article & Videos
Video of the Month: SOUTH KOREA IS OVER
Book Recommendation: Outliers: The Story of Success
Article: I didn’t read one this month, please send me if you read good articles.
Last words
Thank you for reading my first newsletter to the end. It may not yet meet the quality I envisioned, but I’ll keep working on it. Your support means a lot—whether through reviewing it or sharing ‘A Generalist’ with your friends. I deeply value your feedback, so please don’t hesitate to reach out to me on X or via email.
Bonus
In this first issue, I'm very excited to give away some gifts. You can get a 40% discount on ThinkBuddy (TB) annual plans with the code AGENERALIST40
. TB is my daily AI tool that I can access to most of the LLM models and a lot of extra features. (ThinkBuddy)
And this month I’ll give 5 Manus invites. Anyone interested, just email me. First come, first served. This only for subscribers.