Summary

The 90-minute U.S.-China leaders’ call at least put the tariff war back within the scope of negotiations. Trump singled out rare earths, leading 李丰 to infer that China’s export controls may have created more leverage than expected. The timing—around 7 p.m. Beijing time, relatively early in the U.S. day—suggests Washington was more proactive. The two sides did not discuss Ukraine or Iran, but did discuss rare earths, Taiwan and visas for Chinese students. Mutual invitations for the presidents and their wives to visit suggest the agenda could cover purchases, energy, eased rare-earth controls, investment and tariffs, though Trump’s reversals remain the biggest variable.
The problem with China’s May data was not that orders disappeared entirely, but that the tariff war “interrupted expectations”: even with orders in hand, companies are reluctant to expand capacity, build inventory or hire. Official PMI data show a stronger recovery among large firms, a rebound among small and micro firms from April, and weakness among midsize firms. 李丰 attributes the gap to customer mix and sampling methodology. Per-capita spending during the Dragon Boat Festival was nearly flat year on year. If negotiations can contain volatility before July, expectations could rebound from a trough—and “it matters more to China than to the U.S.”
Supply chains are not simply exiting China; they are forming a two-system structure divided by market: U.S.-bound orders are shifting to India and Southeast Asia, while China-based supply chains serve China, Europe and other markets. Apple is the clearest example, with October-November deliveries of its new models set to test India’s assembly capabilities. AI companies could adopt a similar structure, keeping U.S. operations in Silicon Valley or Singapore and China and global operations in mainland China or Hong Kong. Over the long run, companies will still reduce their reliance on any single overseas market.
The Musk-Trump rupture has forced Tesla and SpaceX to be repriced as assets with significant political dependence and key-person risk. Tesla has already gone through two rounds of politically driven volatility: rising after Musk campaigned for Trump, falling amid the Department of Government Efficiency backlash, recovering as Musk stepped back, and dropping again after their public feud. SpaceX remains highly dependent on NASA missions and government contracts. Musk initially only said the tax bill’s “big and beautiful” ambitions were difficult to reconcile, but the dispute quickly escalated into contracts, immigration status and background checks.
Trump is trying to recreate Reagan’s tax-cutting, pro-business and MAGA narrative, but lacks the external tailwinds of globalization, financial liberalization and subdued inflation that Reagan enjoyed. The hosts are more concerned that his decisions follow only a “first derivative”: he sees a problem, picks the most direct Plan A, and fails to fully assess the consequences if it succeeds, fails or is only half implemented. Short-form social media amplifies this simple, forceful style of expression.
The U.S. economy may be entering a service-sector inflation chain of “plenty of jobs, fewer cheap workers, rising wages, but still not enough hires.” The latest data discussed show job creation far below expectations, continued labor demand and hourly wage growth of about 4%. Tighter immigration and other economic policies could first push up wages in restaurants and similar industries, then flow through to consumer prices. With services accounting for more than 80% of U.S. GDP, simultaneous weakness in tourism, financial markets and America’s international image could produce a more difficult mix of slower growth and higher inflation.
Trump’s support for controlled nuclear fusion looks more like an attempt to bypass new-energy fields where China has a strong lead and compete directly for the next energy source of the AI era than evidence that a technological breakthrough is imminent. 李丰 remains highly cautious: “Controlled nuclear fusion is mainly a technology problem,” and whether the policy survives changes of administration is equally uncertain. But if the U.S. keeps adding support, China will likely follow, because mastering fusion means “becoming the sun ourselves”—a metaphor with philosophical, not engineering, force.
Stablecoins are currently competing over whether the dollar or the Hong Kong dollar becomes the anchor of the digital world, but the long-term prize is whether data ownership, tracking, pricing and revenue sharing can open the crypto economy’s “open loop.” Wallets, exchanges, collateral and leverage remain an internal closed loop for digital currencies. AI’s demand for non-public data will force the market to answer whose data was used, how much value it created and what should be paid. The primary market has only limited consensus: biotech has bottomed, robotics remains hot, DeepSeek is driving greater diversity in compute architectures, and most of the AI projects discussed on the program remain “human in the loop.”

Deep dive

1. The U.S.-China Leaders’ Call Exposed Rare Earths’ Real Leverage

李丰 worked backward from the release schedule: Xinhua issued a one-line bulletin at 20:50 Beijing time, followed two hours later by a detailed Chinese-language report. If the call lasted 90 minutes, it may have begun around 7 p.m. Beijing time, relatively early in the U.S. day. He cautiously inferred that “the U.S. was probably still the side pushing for the call,” while stressing that this was only a reading of the details.
Neither side’s public account mentioned Iran or Ukraine, yet Trump singled out rare earths. Someone in the program’s CEO group chat commented, “It looks like rare earths are having a bigger impact than we imagined.” 李丰 said that after months of tight controls, rare earths could affect motors, automotive supply chains and some military equipment, creating a genuine choke point.
The Chinese-language account presented an asymmetric set of concerns: rare earths were a commercial supply-chain issue for the U.S., while Taiwan was a red line China reiterated; visas for Chinese students were also discussed. 李丰’s summary was that the two sharpest issues—Taiwan and rare earths—at least received some interim handling in this call.

2. Trump Is Still Using the Old “Pressure First, Negotiate Later” Script

李丰 linked the recent restrictions on EDA chip-design software, aviation-engine exports, tighter U.S. visa rules and Taiwan rhetoric at the defense ministers’ meeting into a single pattern. It resembles Trump’s standard playbook from his first term: “Every time negotiations are about to start, he begins collecting leverage.”
The pattern also offers a forward-looking signal: if another string of temporary pressure measures targeting China suddenly appears, it may not mean negotiations have ended. It could instead mean Washington is preparing to “negotiate another round of something.” The method itself, however, continues to generate uncertainty and volatility.
Looking back at reports in March that the two leaders might meet in April, 李丰 speculated that the meeting may have collapsed because key conditions could not be agreed, eventually producing the April 2 tariff policy instead. He guessed the sticking points might have included Trump’s proposed tariffs, 50% of which would target China, or other leverage the U.S. had not secured—but explicitly labeled this as speculation.

3. Mutual Invitations Put the Tariff War Inside a Negotiable Framework

The Chinese account mentioned inviting the U.S. president and his wife to visit China at an appropriate time, while Trump responded in noticeably more courteous and enthusiastic language about reciprocal visits. 李丰 said that if the purpose were merely to exchange views on the situation, there would be no need to go as far as mutual invitations. The signal at least suggests that both sides are moving toward a view that negotiations have a defined scope and framework.
The precise exchange remains impossible to determine. During a U.S. visit to China, Beijing might increase orders for Boeing and energy products or partially ease rare-earth controls. During a Chinese visit to the U.S., the agenda could include investment, tariffs, trade, and even AI and chips. “It will be one, two or three of these things—nothing more than that.”
The program did not equate the invitations with a deal already being reached; it limited the conclusion to “negotiable.” Trump could change his messaging at any time, but with only about a month left before July and the tariff pause expires, a further sharp reduction in volatility would be particularly important for repairing Chinese companies’ expectations.

4. Ukraine Has Been Temporarily Downgraded as Washington Has Already Secured Some Interests

Trump explicitly said Ukraine was not discussed on the call. After Ukraine attacked Russian airfields, his public line shifted toward the view that Russia would inevitably retaliate and that the two sides could “fight it out for a while, then negotiate,” possibly with sanctions imposed on both Russia and Ukraine. 李丰 therefore judged that Ukraine had been set aside temporarily rather than headed for an immediate ceasefire.
The program cited Reuters and Bloomberg reports depicting a negotiating deadlock: Russia’s demands are extremely hard-line and largely unacceptable to Ukraine. As the situation intensified, relevant expatriates, students and other personnel also received risk alerts.
From Trump’s perspective, 李丰 argued that the Ukraine minerals agreement had already locked in “at least more than half” of Washington’s objectives. What remains is when those gains will be realized after a ceasefire, not whether the interests were secured. Trump may therefore care less about whether the two sides stop fighting immediately or continue for some time.

5. PMI Shows Orders Recovering, but Business Confidence Has Not

Dragon Boat Festival data extended the pattern of a weak recovery: domestic tourism increased slightly, while per-capita spending was almost unchanged year on year. Shanghai performed relatively well and Beijing was acceptable, but the national total was broadly in line with last year. Uneven performance did not change the picture of a flat aggregate.
Official PMI and Caixin PMI diverged again. Based on the breakdown by firm size, 李丰 inferred that large companies in the official sample were recovering relatively well, small and micro firms had improved from the early-April trade-war shock, and midsize firms were middling. Caixin may include more small and medium-sized companies, which could explain the difference, though he explicitly described this as a guess about sample composition.
Large companies can recover through scaled customers, while small firms may pick up rush orders or temporary orders. Midsize companies depend on several fixed customers but are also only one of many temporary suppliers, leaving them exposed to volatility at both ends. This customer structure is 李丰’s central explanation for their weaker performance.
The more important signals come from medium-term indicators such as inventories, hiring and willingness to produce. New orders themselves are still acceptable, but even companies receiving orders are unwilling to expand capacity, build inventory or hire. The real damage from the tariff war is that it “interrupted expectations,” not that it simply reduced current orders to zero.

6. Supply-Chain Restructuring Will Follow Sales Markets, Not Move Entirely

李翔 argued that even if negotiations stabilize, individual companies will still reduce their reliance on foreign trade. 李丰 agreed, describing the long-term direction as splitting supply chains by end market rather than using attention-grabbing claims such as “Apple’s entire supply chain is moving to India.”
For Apple, Southeast Asia may produce key precision components for the U.S. market, with India handling assembly. China, Europe and other markets will continue to rely mainly on a China-centered supply chain. Deliveries after the launch of new models in October-November will provide a test of India’s supply-chain capabilities.
AI companies could form a similar structure. Companies serving only the U.S. market could move to Silicon Valley; teams serving the U.S., China and other markets could split into two companies, with one based in mainland China or Hong Kong and the other in Singapore or elsewhere handling U.S.-related data and AI businesses.

7. The Musk-Trump Alliance Contained a Control Fight from the Start

From Musk’s explicit support for Trump last August, through his high-profile role in DOGE, gradual withdrawal, public opposition to the bill and eventual feud, Tesla’s stock went through two rounds of politically driven volatility: up after Trump’s election victory, down amid backlash to Musk’s government role, back up as he stepped away, and sharply lower again after the rupture.
李翔 said Musk’s shift toward the Republican Party was not sudden. His opposition to “woke culture,” family experiences and a worldview centered on changing the world overlap with parts of today’s Republican platform. At the same time, his electric-vehicle and rocket businesses depend heavily on government orders, tax incentives and policy, making a slide into politics almost natural.
The fundamental conflict was not just ideological; it was a contest for control between two enormous egos. 李翔 noted that after guiding Tesla through its most difficult period, Musk had rarely worked again in an environment where he lacked absolute control. His clashes with the Treasury secretary and other officials after entering government also meant he was competing over who had greater influence on Trump.

8. Political Dependence Is Re-entering Tesla and SpaceX Valuations

Musk initially expressed himself cautiously: he still supported the president but believed the tax bill’s “big and beautiful” ambitions were difficult to combine. The White House said Trump had known about his concerns and rejected them. The dispute then escalated step by step into government contracts, immigration status, background checks and Musk’s government service.
Another earlier warning sign was Trump’s removal of the NASA administrator nominee recommended by Musk. As the program recalled, SpaceX’s survival from inception through its early years depended heavily on NASA’s continued outsourcing of missions. Even after repeated launch failures, the company continued to receive opportunities and finally succeeded when its funding was nearly exhausted. Government ties were not an external variable; they were part of the company’s growth story.
For a company that 李翔 believes may already be valued at “hundreds of billions of dollars,” the political rupture is not merely media noise. Tesla faces consumer and shareholder sentiment, while SpaceX faces direct mission, contract and regulatory risks. Musk’s personal political behavior has become a shared key-person risk for both companies.

9. Trump Wants to Learn from Reagan, but Today Lacks Reagan-Era Tailwinds

李丰 understood Trump’s initial path as an attempt to recreate Reagan: MAGA, tax cuts, and favoritism toward large companies and high-net-worth groups, with the belief that businesses and the wealthy would convert their capabilities into growth and employment. 李翔 added the Laffer curve and “trickle-down economics,” while noting that the mechanism by which tax cuts later increase fiscal revenue has remained controversial.
One major difference is interest rates. Volcker’s aggressive tightening continued into the early part of Reagan’s term, followed by a brief recession and then recovery. 李丰 also recalled that debt burdens and the share of Treasury debt held by the public rose significantly during Reagan’s presidency, with defense spending an important factor.
Reagan also benefited from supply-chain globalization, cross-border capital flows, floating exchange rates, a stronger dollar and the Plaza Accord, allowing the U.S. to draw on global production, consumption and financial markets. Trump faces a completely different globalized environment and a society less willing to tolerate widening inequality.
One counterintuitive historical observation is that Republicans have long emphasized tax cuts and small government, yet fiscal deficits have often expanded under Republican presidents. The two hosts noted that from the 1980s to today, the only administration to turn a deficit into a surplus was Clinton’s.

10. “First-Derivative” Decision-Making Is the Root of Trump’s Policy Volatility

李丰 gave Trump’s decision-making style a vivid definition: it has only a “first derivative.” Once Trump sees a problem and identifies Plan A, he pushes it forcefully without sufficiently considering what happens if A succeeds, fails, is only half implemented or creates additional problems.
李翔 cited a first-term example from a journalist who became famous for his Watergate reporting. Staff prepared Plans A, B and C for an overseas military operation, intending to show the president two extreme options and steer him toward the middle. Trump saw the most aggressive Plan A, ignored the other two and said, “Good, let’s do this one.”
The two hosts used Reagan as a counterexample. Though not a traditional political figure, he relied more heavily on processes in which advisers reached agreement beforehand. During one discussion with the Israeli prime minister about the Israeli-Palestinian conflict, Reagan took out the paper his staff had prepared and answered strictly according to the agreed position. 李翔 concluded that Reagan at least had less of an ego on critical issues.

11. Short-Form Social Media Is Political Infrastructure for Trump, Not Just a Distribution Channel

The program connected presidential politics to changes in media: Roosevelt in the radio era, Reagan in the television era and Trump in the social-media era were all shaped by the dominant medium of their time.
Truth Social and X both favor short posts, allowing Trump to deliver forceful messages in one or two sentences without constructing a complete or complex argument. 李丰 even judged that if the country returned to an environment dominated by televised debates or long-form internet essays 20 years ago, this communication style might not work in Trump’s favor.
李翔 added that the current campaign had fragmented media channels. Harris more often accepted CNN-style traditional television interviews, while the Republican camp moved into podcasts and YouTube livestreams, then criticized television interviews as “rehearsed in advance and edited afterward.” The ability to sound like a normal person in conversation became a political selling point in itself.

12. Low Tolerance for Dissent Is Becoming a Governance and Asset Risk

The Musk episode led 李丰 to see a broader signal: in America’s public debate, it may be becoming increasingly difficult to discuss the president and his policies from different perspectives—“you can only broadcast one voice.” Musk initially did not call for impeachment; he merely expressed reservations about the bill, yet quickly faced political retaliation.
The Federal Reserve chair offered a contrasting response. He repeatedly said he would watch the data, employment and inflation, without directly saying Trump was wrong, did not understand economics or that he disagreed. The program viewed this caution as evidence of the risks attached to public dissent.
For investors, this is not merely a values debate. Once disagreement can trigger consequences involving contracts, regulation, immigration status and background checks, the relationship between companies and government moves beyond ordinary policy risk into a constraint on executives’ public speech.

13. Targeting Harvard Maximized the Drama—and Picked the Hardest Hill to Take

The conflict between Harvard and Trump was understood as a symbol of the vast divide between intellectual elites and blue-collar voters. Harvard had not publicly declared support for the students in question; it simply had not expelled them immediately like Columbia University. That itself became part of the conflict.
李丰 believes Trump chose Harvard because of its long history, large student and alumni base, and extensive representation among senior figures in politics and business. It is also not necessarily the school with the highest share of international students or the greatest dependence on international tuition.
李翔’s inference was that Columbia and Berkeley may be more representative of the liberal tradition, but lack Harvard’s communications impact. The hosts compared it to mountaineering: “You only get to post on social media if you climb Everest.” The problem is that the most dramatic opponent also possesses the deepest, most entrenched resources for resistance.

14. U.S. Labor Shortages May Push Up Wages Before Service Prices

The program first reviewed the post-pandemic growth relay. From 2020 through the first part of 2022, direct cash transfers to households drove spending on necessities and lifted financial assets. Biden then used federal-budget subsidies for education, healthcare and caregiving systems, supporting service-sector hiring and high employment in 2023-2024.
After Trump tightens immigration and implements other economic policies, low-wage service jobs may be the first to face shortages. The program’s reading of the latest ADP data was that hiring demand remains high, actual job creation fell to a level far below expectations, and hourly wages rose about 4%—“you want to hire people, but you can’t find the cheap workers you used to.”
李丰 used restaurant dishwashers to explain the transmission mechanism. As lower-cost workers of Mexican and other backgrounds become scarcer, owners must raise hourly wages to attract others. Even if higher pay still fails to fill the jobs, costs will gradually pass through to menu prices. This is the first link in a service-sector price spiral; whether a second link forms will require another 1-3 months of observation.
External demand could add further pressure. Financial assets are no longer rising in a one-way boom, while entry restrictions and changes in America’s international image could affect tourism and spending by foreign capital. With services accounting for more than 80% of U.S. GDP, the interaction among labor, tourism and financial channels warrants continued monitoring.

15. China’s External Expectations Spent Two Years Bottoming, Making Stability More Valuable

李丰 placed the turning point in China’s external expectations in 2022. The Ukraine war, G7 rhetoric, lockdowns in Shanghai and elsewhere, and overseas speculation around the 20th Party Congress combined to drive global impressions, expectations and narratives about China steadily lower.
His timeline has expectations roughly bottoming in late 2022, moving sideways near the bottom through 2023 and 2024, and seeing a partial reversal only around the end of the third quarter of 2024, near “9/24.” The subject here is not China’s economy itself, but how the world understands and prices China.
A sudden escalation in the tariff war could therefore interrupt expectations that have only just begun to recover. If the leaders’ call can keep the coming month within a negotiable range, even without an immediate agreement, it could generate a bottoming-out rebound in confidence.

16. The Middle East Could Remain Violent, but Washington May Not Allow a Full Regional Spillover

李翔 identified three lines of deterioration: the second round of Israel-Hamas negotiations is difficult, Israeli operations are intensifying, the U.S. has again vetoed moves related to recognition of Palestine, and talks over Iran’s nuclear program also appear to be failing.
李翔’s short judgment was that the U.S. probably would not allow the situation to escalate into a broader regional war. 李丰 broadly agreed, while stressing that “the Middle East problem does not depend on the Middle East itself”; U.S. intentions are only one of many variables.
The program therefore offered no firm prediction of a ceasefire or escalation. It focused instead on whether the U.S. will permit further spillover and whether the Iran nuclear and Israeli-Palestinian negotiating tracks fail simultaneously.

17. The Probability of Direct Conflict over Taiwan Has Fallen as the U.S. Faces a Larger Reputational Bet

李翔 believes that China’s performance in several real-world combat scenarios, together with its own nonmilitary tests and launches, has given the U.S. both continued uncertainty and a possible reason to reassess China’s military and manufacturing capabilities. 李丰 agreed. That uncertainty itself raises the threshold for direct intervention.
The key issue is not whether the U.S. remains stronger overall, but the asymmetric wager involved in a direct conflict. If China merely achieves “greater than or equal to,” or if the U.S. fails to win clearly, Washington’s position as the world’s leading military power—built over a century—would be damaged. Its global standing and military-backed influence could suffer far more than China’s.
李翔 added a more realistic and colder speculation: some people may ask whether the U.S. already regards Taiwan as belonging to China. He noted that similar views have been inferred from documents such as the Shanghai Communiqué, though 李丰 said he had not noticed that interpretation.
The hosts also revisited the many predictions made around them in 2022 that claimed certain events would happen in a specific month of a specific year. Many people have now forgotten that they made those calls. The shift shows why geopolitical judgments cannot permanently extrapolate the overwhelming U.S. advantage perceived at a particular moment.

18. Controlled Nuclear Fusion Is a Generational Bet on AI Energy, Not a Near-Term Capacity Story

After Trump signed an order supporting controlled nuclear fusion, related U.S. assets rose briefly. 李翔’s first reaction remained: “Controlled nuclear fusion is mainly a technology problem.” Whether it can be achieved depends on engineering breakthroughs, not political will.
李丰’s policy reading is that the U.S. may recognize that fossil fuels cannot necessarily support the enormous power demand of AI’s future, while it would struggle to catch China head-on in new-energy fields where China is strong. Washington may therefore skip wind and solar and bet directly on the next generation of energy.
Policy continuity is questionable. Trump is simultaneously trying to withdraw some subsidy commitments under Biden’s CHIPS Act, suggesting that long-term technology programs could swing with changes in party control. If the U.S. genuinely continues to increase support, 李丰 expects China to do the same, potentially accelerating the maturation of the technology path.
李丰 closed with a metaphor explicitly framed as philosophical rather than an engineering forecast. Bioenergy, coal, oil, wind and water can all be viewed as different conversions of solar energy, while the sun itself is nuclear fusion. If humans achieve controlled fusion, “we will become the sun ourselves”—perhaps part of the same civilizational phase as truly escaping the constraints of the solar system.

19. Stablecoins Are Competing to Become the “Gold” of the Digital World

The U.S. and Hong Kong announced stablecoin policies within roughly two weeks of each other. The program described stablecoins as a convertible bridge between sovereign currencies and digital currencies: an issuer receives dollars or Hong Kong dollars and issues digital certificates backed by the corresponding assets, rather than creating tokens from nothing without reserves.
For the U.S., control of the stablecoin narrative would allow the dollar to remain the primary anchor for digital-asset trading. Hong Kong is issuing a Hong Kong dollar stablecoin, while the Hong Kong dollar itself is currently pegged to the U.S. dollar. 李丰’s metaphor was that the competition will determine “who becomes the gold of digital currency.”
Hong Kong’s greater significance is its role as China’s offshore financial center. 李丰 said he had not checked the data but still judged that Hong Kong’s IPO volume and value in the first six months of the year may have ranked first globally. Meanwhile, Web3 and digital-currency teams that clustered in Singapore around 2022 have begun moving toward Hong Kong.
This does not mean mainland China is broadly opening up digital currencies. It means experimental opening and related financial policies will be placed first in Hong Kong. 李丰 believes China wants to compete for the digital-currency anchoring position, but will not yet open digital currencies themselves on the mainland.

20. Crypto Has Had No Closed-Loop Breakthrough in 12 Years; AI Data Could Open the Loop

Looking back on his experience as an angel investor in Coinbase and Ripple in 2012, 李丰 said that 12 years later the proven money-making models remain mainly wallets, exchanges, and secondary-market financial services such as lending, collateral and leverage. All of them operate inside a closed loop of digital-currency speculation and trading.
An “open loop” means digital currencies or their underlying ledger technology entering non-crypto settings to support real commerce, payments, lending or other exchanges of value. Blockchain’s burden is to record every transaction, every step and every confirmation. In digital-to-digital transactions, however, the objects being exchanged are themselves digital, so processing efficiency matches the nature of the assets.
Stablecoins are therefore currently competing over the sovereign-currency anchor and financial influence. The long-term winner may instead be determined by “who first solves applications outside the closed loop.” Whoever connects the ledger mechanism to real assets and services will have the larger opportunity in digital-currency applications.

21. Data Ownership and AI Applications Are the Primary-Market Battleground Where Consensus Has Yet to Form

AI makes the open-loop problem urgent. The program said OpenAI has acknowledged that publicly available training datasets are nearing exhaustion, meaning the next step will require more private, non-public and high-value data. That raises questions over per-use charges, pre-purchase arrangements or value-based sharing, as well as how to establish ownership, track usage, record transactions and ensure non-repudiation among all parties.
李丰 sees data as a potential factor of production “far larger than real estate.” China established a Big Data Bureau in 2023, and Shanghai has experimented with collecting and anonymizing public-infrastructure data, which may give it an advantage in coordination. But who can use the data, how much they can use, how it should be priced and how providers should be compensated remain unanswered.
李翔 mentioned the Tencent Yuanbao user agreement, which had generated substantial discussion. As he remembered it, the agreement broadly assigned rights to content uploaded or generated through user interaction to the platform. 李丰 instead emphasized that individuals today cannot know “who used it, what they did with it or how much they gained.” Insurers rewarding step counts while users can cheat by attaching devices to pets shows how crude identity verification and data tracking remain.
The primary market has reached only partial consensus. Biotech bottomed in the first quarter after nearly 3 years of weakness since 2022, AI robotics remains hot, and DeepSeek showed the market that underlying compute architectures may include more than Nvidia or GPUs, with inference creating different hardware needs. The application layer remains a field of competing approaches: search, AI coding, games and entertainment have some early consensus, while the projects discussed on the program are mostly still “human in the loop”—closer to autonomous-driving L2-L3 than a direct leap to L4-L5.

Why We Are Bullish Fiverr CEO & Founder, Micha Kaufman: "If You’re Not Adapting to AI, F* You. You’re Done!"