The Two Biggest Stories of the Year: AI & Tariffs

Key Concepts

• Rare Earth Metals: Essential for modern technology and defense, with China dominating production and refining.
• VANC REMX ETF: An investment vehicle for the rare earth and strategic metals sector.
• RAMP: A financial operations platform that automates business expenses and provides data insights.
• RAMP AI Index: A proprietary index tracking business spending on AI, offering insights into adoption and growth.
• AI Adoption: The extent to which businesses are integrating and utilizing AI technologies.
• Government AI Adoption Survey: A survey-based metric that the transcript argues undercounts AI adoption due to its phrasing.
• Business Spend Data: Transaction-level data from RAMP's platform, used to track AI spending and adoption.
• Tariffs: Taxes on imported goods, and their impact on business costs and consumer prices.
• 996 Phenomenon: A work culture characterized by long hours (9 AM to 9 PM, 6 days a week), observed in San Francisco workers.
• Shrinkflation: The practice of reducing product size or quantity while maintaining or increasing price.

AI Adoption and Business Spending

The discussion centers on the current state of Artificial Intelligence (AI) adoption within businesses, challenging the notion of a potential bubble by examining actual spending patterns. Ara Carzian, an economist at RAMP, highlights the unique data RAMP possesses through its financial operations platform, which allows for granular tracking of business expenditures, including AI.

Key Points:

• Data Limitations: Traditional data sources for AI adoption are scarce, as many AI companies are private, and public information is often biased.
• RAMP's Advantage: RAMP's platform provides line-item receipt data, enabling detailed tracking of AI spending, including API usage, contract renewals, and the adoption of specific AI companies like OpenAI and Anthropic.
• RAMP AI Index: This index reveals that AI product retention rates have significantly increased from approximately 50% in 2022 to 80% in 2024. AI contract sizes are also growing, with an estimated average of $1 million next year.
• Enterprise vs. SMB Adoption: While enterprise adoption of AI has been slower due to concerns about trust and liability, overall adoption by companies on RAMP's platform has reached about 40% with significant AI contracts. This includes spending on platform providers, AI agents, and specialized tools for specific teams (e.g., finance, software engineering).
• Discrepancy with Government Data: RAMP's estimated AI adoption rate of 44% starkly contrasts with a government survey estimate of 9%. Carzian attributes this discrepancy to the government survey's outdated and narrowly defined question: "have you used AI to produce goods and services?" This phrasing, he argues, fails to capture the widespread use of AI in tasks like customer service automation and software development.
• Underreporting by RAMP: Even RAMP's data may slightly underreport adoption due to the increasing availability of free AI options (e.g., Gemini in Google Workspace, free ChatGPT usage).
• Long-Term Outlook: Carzian is optimistic about AI adoption, predicting it will eventually reach near 100% across businesses, even if not always visible to consumers. He anticipates growth in verticalized AI solutions as products mature.

Supporting Evidence:

• Nvidia crossing $5 trillion in market cap.
• RAMP's data showing 80% AI product retention in 2024, up from 50% in 2022.
• Estimated average AI contract size of $1 million for next year.
• 40% of companies on RAMP's platform having significant AI contracts.
• Comparison of RAMP's 44% adoption rate with the government's 9% estimate.

Tariffs and Their Economic Impact

The discussion shifts to the complexities of tariffs and why their anticipated inflationary and growth-inhibiting effects have not been as pronounced as expected.

Key Points:

• RAMP's Tariff Data: RAMP can track tariffs through invoice line items for imported manufacturing goods and import/export duties on receipts.
• Slower Than Expected Impact: While the share of transactions including tariffs is increasing, especially in manufacturing and retail, the overall tariff rate remains low (around 1.4% on average, with about 3% of invoices showing a tariff transaction). This incidence has doubled from last year but is still considered small given the broad implementation of tariffs.
• Frictions in Implementation: Several factors contribute to the gradual and less impactful rollout of tariffs:
  • Trade Agreements: Existing trade agreements can override or complicate announced tariffs.
  • Logistical Challenges: Ports and customs officials face difficulties in assessing and collecting tariffs, leading to delays and training needs.
  • Reclassification and Evasion: Companies may find legal ways to reclassify products or exploit variations to avoid paying tariffs.
  • Anticipation of Rollbacks: Some companies are delaying significant commerce changes, banking on potential tariff rollbacks.
  • Relocation to Lower-Tariff Nations: Instead of returning production to the US, some companies are moving operations to countries with lower tariffs.
• Disruptions vs. Price Increases: While significant disruptions in import/export commerce, delayed shipments, and port operations have occurred, the direct price increases for consumers have been much slower than anticipated because many companies have not yet fully absorbed or passed on the tariff costs.

Supporting Evidence:

• RAMP data showing a slow and granular increase in tariff transactions.
• Average tariff rate of 1.4% and 3% of invoices with tariff transactions.
• Anecdotal evidence of companies moving production to lower-tariff nations.
• Observed delays in shipments and port operations.

The 996 Phenomenon and Labor Market Dynamics

The conversation touches upon the "996 phenomenon" and its potential impact on the labor market, particularly in San Francisco.

Key Points:

• Definition of 996: Working 9 AM to 9 PM, six days a week. This work culture, originally associated with China, is now being observed in San Francisco.
• RAMP Data on 996: RAMP data, specifically takeout invoices, suggests an increase in workers ordering food on Saturdays in the San Francisco Bay Area, indicating they are working more weekends.
• Estimated Impact: An estimated 40,000 workers in the Bay Area may now be working Saturdays and Sundays in a way they weren't last year.
• Broader Labor Market Concerns: The discussion acknowledges the significant question of how AI will affect the labor market. While the optimistic view is that new jobs will be created, there's also concern about displacement in the interim.
• Limited Impact on Most Jobs: Carzian believes that most jobs in the US economy, particularly those requiring a physical component (e.g., nurses, restaurant workers, healthcare workers), are not at immediate risk of automation.
• Current Unemployment: The unemployment rate has not significantly increased over the past two to three years, suggesting that any job displacement from AI is currently occurring in specific sectors and on a narrow segment of workers who are still able to find other employment.

Supporting Evidence:

• RAMP data on increased takeout orders on Saturdays in San Francisco.
• Estimated 40,000 workers in the Bay Area working weekends.
• Observation that most US jobs have a physical component not easily automated.
• Stable unemployment rates over the past 2-3 years.

RAMP's Value Proposition and Future Research

The segment concludes with an explanation of RAMP's core offering and Carzian's research focus.

Key Points:

• RAMP's Core Function: RAMP is a financial operations platform that automates tasks like expense reporting, receipt management, and credit card tracking, freeing up finance teams and individual users.
• Data-Driven Insights: The platform's ability to capture detailed transaction data allows RAMP to provide valuable insights into business spending trends, helping companies make better financial decisions.
• Carzian's Role: As an economist, Carzian leverages this data to identify emerging economic trends, validate or correct public discourse, and provide data-driven perspectives on topics like AI and tariffs.
• Public Accessibility: RAMP's data and Carzian's research are publicly available on RAMP's website (ramp.com/data) and his Substack, "Ramp Economics Lab."
• Future Research: Carzian is open to suggestions for future research topics, aiming to distinguish between popular narratives and actual economic realities.

Notable Quotes:

• Ara Carzian: "When we're talking about whether or not something is a bubble, I feel like people are grasping for all the different data sets that they can possibly use to support their answer. But frankly, for AI, there's just not a lot of data available."
• Ara Carzian: "The real question is, is the technology that being developed providing benefits to the potential buyers such that they'll be likely to buy more in the future and they're actually going to see productivity gains from this."
• Ara Carzian: "The government estimate of AI adoption from US businesses, I do think has one significant flaw. First of all, it's based on a survey and it's literally based on a survey that goes out every two weeks to businesses."
• Ara Carzian: "I see line item receipts which most transaction level data sets just don't have. So that's what allows me to do most of this work."
• Ara Carzian: "I do think AI adoption is going to increase to the point that we're at 100% adoption across businesses."
• Ara Carzian: "If the size of contracts start to go down and if retention rates start to go down that means that companies are trying these AI products and services and they're just not working for them."
• Ara Carzian: "The companies that are producing this technology are producing technology that people are buying. A lot of times they're profitable. Yeah. Um and um they're revenue generating. That was not necessarily the case in the internet bubble."
• Ara Carzian: "We found was that tariffs are increasing like the share of transactions that include tariffs are certainly increasing especially in manufacturing and retail but it's much more slower and more granular than it than you'd think."
• Ara Carzian: "The number of of workers uh at least in the SF area working on weekends like Saturdays and Sundays has increased. We estimate over the entire Bay Area maybe about 40,000 workers across the metro area are now working Saturdays and Sundays in a way that they were not last year."
• Ara Carzian: "Let AI read your receipts and file your finances for you. Let AI process through your, you know, what you're allowed to buy and purchase on an offsite or what you're allowed to do for Door Dash. You don't have to think about that stuff. uh automation of all your finance tasks is really the shorthand for it."

Synthesis and Conclusion

The conversation with Ara Carzian provides a data-driven perspective on critical economic trends, particularly AI adoption and the impact of tariffs. RAMP's unique position, with access to granular business spending data, allows for a more nuanced understanding than traditional surveys or anecdotal evidence. The data suggests that AI adoption is robust and growing, with high retention rates and increasing contract sizes, challenging the "bubble" narrative. The impact of tariffs, while present, is proving to be more gradual and complex than initially anticipated due to various implementation frictions. Furthermore, the discussion touches upon the evolving labor market, with AI's impact being concentrated in specific sectors, and the emerging 996 work culture in San Francisco, evidenced by increased weekend work. RAMP's platform and Carzian's research aim to demystify these trends by providing accessible, data-backed insights for businesses and the public.