Gas Is Back Above $4 — And Could Keep Rising

SUMMARY

Ed Elson hosts a discussion with Matt Smith and Charlie O’Neill, examining the impact of intensifying conflict in the Middle East on oil and gas prices, and exploring the implications of China’s new open-source AI model, Kimi K3. The episode covers the mechanics behind rising fuel costs, the strategic maneuvers in oil supply routes, and the growing influence of open-source AI models on global technology markets.

MAIN POINTS

• Escalating conflict in the Middle East, including US airstrikes on Iran and a new Houthi blockade, has driven oil prices to $89 a barrel and pushed US gasoline above $4 per gallon. [1:30]
• Iran is leveraging its strategic position by threatening to block alternative oil supply routes, intensifying market fears and supporting higher prices. [6:01]
• Despite rerouting efforts, tanker traffic through the Strait of Hormuz remains severely restricted, and oil prices are likely to rise further if the current situation persists. [11:01]
• Chinese startup Moonshot AI releases Kimi K3, the world’s largest open-source AI model, outperforming some Western models and significantly undercutting their costs. [17:01]
• Open-source AI models are increasingly attractive to businesses due to lower costs and the ability to customize, challenging the dominance and margins of closed-source labs like OpenAI and Anthropic. [23:00]
• The AI ecosystem is expected to bifurcate, with frontier labs focusing on cutting-edge research while open-source models serve most economically valuable tasks, leading to broader competition and lower margins. [29:00]
• The episode concludes with an invitation to a live stream discussing the global economic outlook and China’s role in AI, highlighting ongoing analysis of these critical issues. [31:02]

DETAILED ANALYSIS

The episode opens with a sharp focus on the recent surge in US gasoline prices, now averaging above $4 per gallon, and diesel prices exceeding $5, reflecting a 15% increase in just the past week. This spike is directly linked to escalating conflict in the Middle East, particularly the Strait of Hormuz, a critical chokepoint for global oil shipments. The US has conducted airstrikes on Iran for nine consecutive nights in retaliation for Iranian attacks on oil tankers, while Iran’s Houthi allies in Yemen have declared a naval blockade against Saudi Arabia.

This blockade threatens the Saudi pipeline to the Red Sea, previously a vital workaround to bypass the Strait of Hormuz.

Matt Smith, Director of Commodity Research at Kpler, explains that tanker traffic has become increasingly difficult to track due to shifting routes. Previously, Saudi Arabia managed to reroute about half its crude exports—approximately 3.5 million barrels per day—through the Red Sea, cushioning the supply shock for major importers like India, China, and South Korea. However, the new Houthi blockade threatens this alternative, raising the risk of a significant supply disruption.

While the blockade is not yet fully enforced, the mere threat has already had a bullish impact on oil prices.

Smith notes that the dissolution of a memorandum of understanding between the US and Iran has led to a full-scale escalation, with both sides targeting infrastructure and shipping. Despite the high oil price, the most acute pain is being felt in refined products such as gasoline and diesel, rather than crude itself. This is partly because China has sharply reduced its oil imports—by about 5 to 5.5 million barrels per day—helping to offset Middle Eastern production losses.

Additionally, many refineries have scaled back operations, further limiting product supply and driving up prices for end consumers.

Investor sentiment in the oil market is described as cautious, with many traders reluctant to take positions due to the unpredictability of geopolitical developments. The lack of liquidity has contributed to price volatility. The US administration remains highly focused on gasoline prices, but less so on diesel, which has seen even more dramatic price increases.

Looking ahead, Smith warns that if the current standoff continues for several more months, oil prices could easily surpass $100 per barrel. The only scenario in which prices would not rise further would involve significant drawdowns in global inventories and creative rerouting of shipments, both of which appear increasingly difficult under current conditions. The ongoing tit-for-tat attacks and the threat of further blockades suggest that high prices at the pump are likely to persist, with material consequences for US consumers and the broader economy.

The second half of the episode shifts to the technology sector, where Chinese startup Moonshot AI has released Kimi K3, the largest open-source AI model to date. Kimi K3 reportedly outperforms leading models from OpenAI and Anthropic on several benchmarks, including front-end coding, while operating at roughly a third of the cost. This development has immediate market repercussions, with the NASDAQ falling as US tech stocks react to the competitive threat posed by Chinese open-source models.

Charlie O’Neill, Co-Head of Model Training at Baseten, emphasizes that the real story is the rise of open-source AI rather than a simple China-versus-America narrative. Open-source models, including Kimi K3 and others like GLM and Inkling, are rapidly closing the gap with closed-source offerings. The key distinction is that open-source models allow users to download and modify the underlying weights, enabling customization and further training for specific tasks.

This flexibility, combined with lower costs, is increasingly attractive to businesses seeking to integrate AI into their operations.

O’Neill explains that closed-source labs such as OpenAI and Anthropic have justified their approach on safety and security grounds, arguing that only a select few should control advanced AI. However, the economic incentives are significant, with margins rumored to exceed 80% for these labs. The proliferation of large, high-quality open-source models threatens these margins by offering comparable performance at a fraction of the price.

As a result, the AI market is expected to bifurcate: frontier labs will focus on cutting-edge research and specialized applications, while open-source models will dominate the broader market for economically valuable tasks.

This competitive dynamic is likely to result in lower margins for model trainers and a redistribution of profits across the AI ecosystem, benefiting compute and inference providers as well as end users. The episode concludes with a preview of an upcoming live stream featuring further discussion on the economic and technological implications of these trends, particularly China’s expanding role in AI.

LINKS

• Prof G Markets YouTube channel
• Prof G Markets newsletter subscription
• "Notes On Being A Man" book order page
• No Mercy / No Malice newsletter subscription
• Prof G Markets Instagram
• Scott Galloway Instagram
• Ed Elson Instagram
• Ed Elson X (Twitter) profile
• Ed Elson Substack
• Lisa mattress official website
• SoFi Coach financial guidance and sweepstakes

KEYWORDS