May 3, 2025 · 1h 35m · allin
Trump's First 100 Days, Tariffs Impact Trade, AI Agents, Amazon Backs Down
David Sacks · 20m spoken Chamath Palihapitiya · 18m spoken Jason Calacanis · 18m spoken Aaron Levie · 17m spoken Ryan Petersen · 10m spoken
0:00 / 0:00
▶ Watch on YouTube →
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
In this episode of the All-In Podcast, guest panelists Aaron Levie and Ryan Petersen join the hosts to analyze President Donald Trump's first 100 days, trade tariffs, global supply chain disruptions, and media coverage. The panel then transitions into a deep dive on enterprise AI, evaluating autonomous agent workflows, total addressable market expansion, hype cycle realities, and technical scaling bottlenecks.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. The hosts hold 67.8% of the talking time here. How this is scored →
The hosts as informed peer 4.7 Guest teaching 4.4 Guest disagreement 2.9 The hosts pushing back 2.9
speaking balance: gold is the hosts, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 58:40 Aaron rejects historical WTO framing
Aaron forcefully pushes back against Sacks' historical WTO narrative, rejecting 25-year-old context to demand focus on immediate economic friction.
Hardest push from the hosts ▶ 1:03:15 Sacks refuses Aaron's perfect plan framingSacks directly rejects Aaron's critique of the administration's trade execution, asking where his supposed perfect plan was during the previous 25 years.
Biggest teaching moment ▶ 1:09:40 Aaron reframes software TAM expansionAaron educates the hosts on how agentic AI fundamentally shifts software monetization from per-seat user licenses to capturing multi-trillion-dollar labor budgets.
The host holds their own ▶ 1:18:40 Sacks articulates the 3 pillars of AI exponential scalingSacks demonstrates deep technical domain knowledge by detailing how compounding improvements across algorithms, chips, and data center compute yield a million-fold progress vector.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | The hosts as informed peer | Guest teaching | Guest disagreement | The hosts pushing back | Why |
|---|---|---|---|---|---|---|
| Pre-Show Banter: Flight Schedules and Private Aviation | 1 | 1 | 2 | 2 | The hosts and guests engage in playful banter about private jet logistics, missing commercial flights, and online comment section critiques. | |
| David Sachs' Government Role and D.C. Private Club | 2 | 3 | 2 | 3 | Jason probes Sacks on his government time-tracking and reports of a $500,000 private club in D.C. Sacks clarifies that $500k is for founding members only and explains the club's rationale. | |
| Comprehensive Assessment of Trump's First 100 Days | 4 | 2 | 3 | 3 | The panel reviews Trump's first 100 days. Chamath gives a B+ grade citing foreign direct investment and border security, while Aaron praises AI policy but critiques potential tariff headwinds. | |
| Ukraine Conflict, Foreign Policy, and Nuclear De-escalation | 5 | 1 | 1 | 3 | Chamath shares personal insights from a dinner with President Trump regarding nuclear existential risk, while Sacks details the history of U.S. weapon escalations in Ukraine. | |
| Downstream Tariff Impacts on Global Supply Chains | 3 | 7 | 3 | 2 | Ryan Petersen educates the hosts on supply chain mechanics, explaining that ocean freight bookings from China dropped 60% and outlining how bonded warehouses work. | |
| Economic Strategy: Central Planning vs. Free Market Acceleration | 3 | 7 | 6 | 3 | Aaron Levie strongly advocates for acceleration over tariffs, calling to replace Navarro and arguing that tariffs act like socialist central planning that distorts free markets. | |
| Trump's Policy Communication vs. Mainstream Media Coverage | 5 | 4 | 6 | 6 | Chamath argues mainstream media ignores pro-business tax policy additions like factory PPE expensing, while Aaron forcefully replies that administrative instability and rapid updates confuse CEOs. | |
| Amazon's Tariff Notification Flip-Flop and Chinese Seller Dominance | 6 | 5 | 8 | 8 | Debate escalates over Amazon's tariff line items and Chinese third-party seller dominance. Sacks challenges Aaron for presenting late-stage critiques without having proposed solutions during the previous 25 years. | |
| Manus AI Demo and the Agentic User Interface Paradigm | 8 | 2 | 1 | 1 | Sacks breaks down the Manus AI demonstration, highlighting its two-pane UI paradigm and integration with the Model Context Protocol (MCP) standard. | |
| Enterprise AI Agent Adoption and Software TAM Expansion | 4 | 8 | 2 | 2 | Aaron Levie educates the hosts on how AI agents expand software TAM by transitioning pricing models from per-seat licenses to capturing underlying labor spend. | |
| Real-World Logistics Applications and Expanding Work Scope | 3 | 7 | 2 | 1 | Ryan Petersen demonstrates real-world AI voice agents placing automated phone calls to truck drivers, illustrating how low-cost AI creates entirely net-new operational capacity. | |
| The Enterprise Trough of Disillusionment and Hype Cycle Debate | 6 | 4 | 3 | 5 | Chamath highlights the enterprise trough of disillusionment caused by model hallucinations in regulated industries, prompting pushback from Sacks who argues adoption is expanding. | |
| Three Pillars of Exponential AI Scaling and Evaluation Benchmarks | 8 | 6 | 2 | 2 | Sacks lays out a detailed framework on the three compounding pillars of AI scaling (algorithms, chips, and data centers), projecting a million-fold improvement over four years. | |
| Polymarket Predictions and Validation Constraints in AI | 7 | 3 | 1 | 2 | Sacks and Chamath analyze why fields with objective compilers like coding and math progress much faster in AI than subjective fields like legal analysis. | |
| Token Intensity, Customs Classification, and Improvement Engineering | 6 | 6 | 1 | 1 | Ryan explains LLM accuracy in classifying products for customs duties, while Chamath introduces 'improvement engineering' as the essential discipline for reducing enterprise AI error rates. |
Statements from this episode (34)
Calacanis: Uber is an anti-tariff stock unaffected by trade tariffs
“All that matters to me is that Uber is the anti-tariff stock. It just does great. It's not impacted by tariffs.”
Sacks says $500K fee for D.C. club applies only to founding members
“That's true for founding members who have additional benefits, but there's also a lower level. That's the more reasonable membership level.”
Sacks joins Trump Jr.'s D.C. club as first member, avoids ownership
“Since I'm in the government, I can't be an owner, but I told him I'd be happy to be member number one.”
Chamath: Foreign direct investment into the US under Trump exceeds $1T
“I think it's approaching, if not, it has already exceeded a trillion dollars from corporations and organizations and individuals from around the world who have committed to bringing money into the United States.”
Chamath: Trump reduced illegal US border crossings to zero
“I'm just saying getting the illegal crossings to zero and he's done that.”
Sacks: Trump ended use of disparate impact for affirmative action
“He also signed an executive order ending the use of disparate impact for affirmative action.”
Chamath reveals he and Nat Friedman recently dined with President Trump
“Nat and I had dinner with POTUS two weeks ago.”
Chamath: Trump has shown there are no conditions under which he'll go to war
“I think that Trump in the first hundred days, to your point, I think has completely reinforced that there are no conditions under which he'll go to war.”
Sacks: US would be heading to World War III under Democrats
“I have no doubt that if the Democrats were still in office, we would be in an escalatory spiral right now with the destination being World War III.”
Petersen: Ocean freight bookings from China to US dropped 60%
“And what has happened is a 60% decline in bookings of ocean freight from China to the US.”
Petersen: Companies stay in China for quality manufacturing, not cheap labor
“Labor costs in China are not, are you're not there for cheap labor. You're there for quality manufacturing at this point. Like there's much cheaper labor in Southeast Asia or other parts of the world than there is in China.”
Aaron Levie says the Trump administration should fire Peter Navarro
“I'd get rid of Navarro immediately.”
Levie: Government supply chain mandates cause socialist-style economic distortions
“That, that ends up with just lots of economic distortions that everybody on the right would have called, you know, the left socialists. For trying to kind of implement central planning around supply chains.”
Chamath: Bessent says tax bill allows full factory expense deductions
“This past week, Besant said that the tax bill will allow you to fully deduct all the PPE and all of the incidental costs of building a factory.”
Petersen: Trump administration exempted auto parts from tariffs
“But also earlier this week, they made it so that auto parts are not subject to the tariffs. They created a huge exemption.”
Petersen: 60% of Amazon third-party sellers are Chinese companies
“I think it's 60% of all the sellers on Amazon are these Chinese registered. They're not registered in the United States at all, just Chinese companies selling.”
Sacks: China annihilated US rare earth processing capability over 25 years
“China over the last 25 years has been able to strategically annihilate our rare earth processing capability and our ability to cast rare earth magnets.”
Jason Calacanis reports seeing layoffs following economic policy changes
“We've already started to see layoffs.”
Calacanis: OpenAI plans to charge $2k-$20k monthly for AI agents
“Open AI is planning to charge between two and 20 K a month for different levels of AI agencies would be basically cron jobs.”
Sacks says Model Context Protocol standard is rapidly taking off
“And in fact, there's a new standard called MCP, which is taking off like wildfire, which is built specifically to enable agents to connect with applications and understand the data and understand the actions that are possible in those SaaS applications.”
Sacks predicts split-pane agentic interfaces will become standard AI UI
“Look, Manus is just the tip of the iceberg here. I think this will become a very standard UI paradigm.”
Calacanis: Launch built AI agent processing 20,000 startup applications yearly
“We have 20,000 applications a year, and we have updates coming in from investments. We are now taking those sacks, Aaron, and we are having an agent sort them and then look for competitors and compare them to the last update. And we're looking into our notion,…”
Levie: AI agents will massively expand software TAM by capturing labor spend
“You know, I would argue, you know, strongly that's a massive TAM increase because now software starts to go after labor spend.”
Levie: Most AI agent value comes from new capabilities, not worker replacement
“I don't like the word replace because I think actually most of the upside is actually going to be for companies that now deploy labor at things that they wouldn't have deployed labor at before.”
Petersen: Flexport makes thousands of daily AI calls to truck drivers
“So like we, we're making thousands of phone calls a day using AI calling truck drivers.”
Chamath: Class action lawsuits over LLM errors in enterprise are inevitable
“And I, I'm guaranteeing you, we have not seen the class action lawsuits that will come when those errors will eventually be made. They're guaranteed to be made. We just don't know the scope and the scale of them.”
Sacks: Enterprise AI hype cycle has not reached peak expectations yet
“I don't see the disillusionment. I don't know where this is coming from. I don't even think we're at the peak yet.”
Sacks predicts AI advances will yield a 1,000,000x increase in capability
“So you multiply those things together. The algorithms, the chips, and then the raw compute that's available. You're talking about a million X increase, some of which will be captured in price reductions. Some of it will be in the performance ceiling, and then …”
Levie: Best AI models reach 90% accuracy on Box's single-pass document benchmark
“We send it 500 documents, and we ask for 40 data fields back from the documents, and so it has to get every single data field correct, and we only do a single pass, so we send the document to the model, we get a single pass back. Right now, the best score is a…”
Chamath: 8090 uses Gemini for general tasks and Claude for coding
“So for many tasks at 8090, we use Gemini. It's incredible. But for most of our co-gen, we use Anthropic and Claude kicks ass. It's exceptional.”
Sacks: Code compilation enables rapid AI breakthroughs via objective feedback
“In any event, the reason why the progress is so rapid with coding is because code compiles and you can Determine objectively whether it works or not. You can validate it. And so that makes it a perfect area for AI to get better at through reinforcement learnin…”
Sacks: Massive AI CapEx buildout makes sense given token intensity
“So the amount of compute required to serve all these new applications is going to be massive, which is why I think the CapEx build out actually makes sense.”