This week the hosts sit down with Saikat Chakrabarti, who was Stripe’s second engineer and later worked on the Bernie Sanders campaign and served as Alexandria Ocasio-Cortez’s first Chief of Staff. He is now running for California’s 11th House district, which represents San Francisco, against Nancy Pelosi and Scott Wiener. Pelosi has held this seat for nearly four decades.
They talk about why Saikat left tech and how to encourage more young and ambitious people to get involved in politics. Saikat shares his vision for a Democratic Party that supports candidates that are builders from different backgrounds and do not take donor money.
The conversation dives into key policy issues, including:
They also discuss AI regulation. Saikat argues that focusing only on catastrophic AGI risk overlooks the more immediate concern: concentrated economic power and widespread job displacement.
The episode ends with a fun speed round where Saikat responds to his haters including YC CEO Garry Tan.
You can learn more about this platform at https://www.saikat.us/en.
If you want to bet on Saikat, or on the odds of basically anything else, download Kalshi. You will receive a $40 bonus after trading 100 contracts: https://kalshi.com/r/getthecheck
00:00 Intro00:34 From Stripe’s second engineer to a bid for Congress05:31 DOGE09:12 Changing the Democratic Party18:45 Why PG&E shouldn’t exist31:26 The free market and when it fails33:28 How Congressional Reps influence city budgets39:33 Reviewing Daniel Lurie’s first year as mayor42:33 The homelessness problem in SF48:18 AI regulation52:11 Responding to the haters55:32 AOC running for president in 202858:07 How to get involved with Saikat’s campaign
Maya, Priya, and Anika start by unpacking why they need to edit the podcast, while their other podcaster friends just upload the raw file.
This week they dive into reports that the Trump administration may (or may not?) take a stake in quantum computing companies. The hosts break down what quantum is and what the important use cases are. They dive into Google’s new “Quantum Echo” breakthrough, which is basically the first time a quantum computer has done something verifiably useful and repeatable.
Next they talk about the latest X tech drama, which is David Sacks the Crypto Czar arguing with Anthropic about AI regulation like SB 53. The hosts outline SB 53 and discuss whether it actually hurts AI startups like Sacks claims. They also talk about how Trump tried to pass anti-AI regulation in the “Big Beautiful Bill” and what the future of regulation can and should look like.
After a brief tangent on cafes in SF, it’s time for the usual Get the Check segment. The girls investigate a viral video accusing food distributor Sysco of ruining American restaurants. While mostly true, the viral video got some key stats wrong.
00:00 Intro00:30 Rumors Trump admin is taking stakes in quantum computing03:58 What is quantum computing06:46 What use cases does quantum computing solve12:17 Google's Quantum Echo breakthrough16:22 David Sacks vs. Anthropic on AI regulation23:59 Jack Clark’s essay26:57 SB 1047, Safe and Secure Innovation for Frontier AI Act32:11 Our thoughts on the AI regulation debate37:27 State vs. federal AI regulation43:20 Would Sysco “Get the Check”46:09 Side note about coffee shops in SF51:34 Sysco’s company background52:23 Sysco's controversies
It’s day 21 of the government shutdown. The pod’s true fans were there when Maya told everyone to put money on the Kalshi 15+ days shutdown market two weeks ago.If you still haven’t made a Kalshi account, you’re probably not a true fan, but on the upside, if you do it now, Kalshi is giving Get the Check listeners $40 after they trade 100 contracts if you use this link:https://kalshi.com/r/getthecheckThe hosts start by talking about China’s new rare earth mineral export controls. They added another five to the list, so there are 12 now if you’re keeping track like we are. Everyone’s always talking about how cooked the US is if we don’t have rare earths, but the hosts also dive into the history of the market and how China was able to monopolize it. TLDR: we basically let them, but hindsight is always 20/20 (and some people’s foresight is also 20/20, like Maya with the Kalshi markets). The US is now partnering with other countries, and we’re even willing to go Down Under and start a $700M fund with Australia to get our hands on these minerals.Next, the show takes a sci-fi turn with the Anduril EagleEye glasses. The last time the military tried AR/VR glasses, soldiers performed worse with them on, but now there’s a take two, with Microsoft in the backseat and Anduril leading. The goal of EagleEye is to create a digital map of the battlefield so that soldiers get information from other soldiers’ feeds, drones, sensors, and more.Taking a quick intermission to ask you to rate us on Spotify and give us a thumbs up on YouTube. It helps our profitability, which is hanging on by a thread, much like Zepto’s…Zepto is GoPuff for India, and they just raised $450M at a $7B valuation. Speaking of India, Happy Diwali y’all! In India, people go to the grocery store four times more frequently, so Zepto introduced 10 minute delivery, which is the only way delivery can be even quicker than walking a few blocks. They own the full stack of delivery, unlike other apps that deliver goods from third parties like grocery stores. Tune in to find out if Zepto will “Get the Check.”00:00 Kalshi government shutdown market
02:47 China’s rare earth mineral export controls
16:45 Anduril's EagleEye AI helmet
35:14 Zepto raises $450M for delivery in India
It seems like Anika was waiting for this week’s recording to confront Maya about something she said (or rather didn’t say) in the group chat. Once they get that out of the way, the hosts discuss Figure’s 03 humanoid announcement, the Israel-Palestine deal Trump is brokering, and OpenAI’s Dev Day.
Figure 03 is Figure’s latest humanoid and the pod says it would get their check 1000%. The hosts talk about product improvements compared to 02 and how 03 is focusing on a humanoid in the home use case. They also dive into Helix, which is its AI system that allows the humanoid to receive natural language prompts like “go wash the dishes.” AI is a critical breakthrough for humanoids because it allows them to do tasks they haven’t seen before and haven’t been explicitly programmed for. Finally, the hosts talk about the realistic timeline for these getting into homes and commercial spaces.
Next, they cover OpenAI Dev Day. OpenAI might be going from a chat interface to full operating software with in-chat apps like the ability to @Zillow and an apps SDK, which means developers can build in the OpenAI ecosystem. They also cover the AMD deal that pairs 6 GW of planned compute with the option for OpenAI to buy up to 160 million AMD shares at a cent each. The market has been buzzing since, with AMD popping in a good way, but some investors worried AI is going to pop in the bad way.
00:00 Intro
03:14 Figure announces the Figure 03 humanoid
35:43 OpenAI Dev Day and AMD deal
It’s Tech Week and Get the Check’s one-year anniversary! Maya may have never believed in this milestone, but we know you all did. This Tech Week, the pod is co-hosting a DJ set at the Composio office on Sunday morning. Link’s in the bio!
The hosts kick it off with DoorDash’s new delivery robot called Dot. It is an in-house autonomous vehicle built by DoorDash Labs that can drive on both roads and sidewalks. They break down how it could automate up to half of DoorDash’s deliveries, why Phoenix is the testing ground for every AV startup, and how dot’s modular design could make it surprisingly profitable. Priya and Anika also discuss how Berkeley lowkey had food delivery bots five years ago.Next, they dive into the October 1st government shutdown, explaining why Democrats are using the tactic to force Republicans to extend Affordable Care Act subsidies and reverse Medicaid cuts from the “Big Beautiful Bill.” They trace how the shutdown connects back to Trump-era tax cuts, which states have the most to lose, and why Democrats are betting that voters will still blame Republicans despite initiating the standoff. The hosts also discuss the shutdown’s real-world effects like furloughed workers, national parks going unstaffed, and what it means for the market.Finally, they cover Tinker, the first product from Thinking Machine Labs, Mira Murati’s $12 billion AI startup. They explain how Tinker’s fine-tuning API lets researchers train models with reinforcement learning (RLHF) without managing GPUs, which is a move that could make AI development cheaper and more customizable.
00:00 Intro
02:23 DoorDash launches Dot
17:54 Why the government is shut down
30:58 Thinking Labs launches Tinker
The hosts are back and diving into a massive week for OpenAI. They start with Nvidia’s $100B investment, which will fund OpenAI’s GPU purchases, and explain why this circular deal doesn’t necessarily signal an AI bubble. They also break down OpenAI’s trillion-dollar infrastructure buildout and the roles Oracle and CoreWeave are playing.
Next, they cover OpenAI’s new product Pulse and the growing rumors about ads. Maya and Anika debate whether Pulse will actually work. Comment whose side you’re on and if you’ve tried Pulse (or know someone who can plug the pod a Pro subscription).
Then they shift to TikTok’s final deal, announced by Trump, and why it’s such an atypical M&A process. With handpicked buyers instead of an open bidding war, the hosts talk about what this means for creators, the algorithm, and TikTok’s valuation.
Finally, in the Get the Check segment, they cover Until Labs, a biotech startup that just raised $58M to push forward organ cryopreservation. Their long-term vision is extending the life of donor organs…and eventually freezing humans to pause biological time. If you’re wondering how, tune in for the full breakdown.
This week the hosts are back, recharged after a trip to Thailand. They kick things off by recapping their first-ever female founders + investors dinner and share plans to host more events soon. Follow along on socials at @getthecheckpod to hear about the next event.The podcast dives into Meta Connect and the launch of Meta’s Ray-Ban Display and Neural Band. The glasses let you talk to Meta AI to find a coffee shop, get directions, and text a friend - all without picking up your phone. They also break down Meta’s choice to use a neural band instead of eye tracking, which enables gestures like swiping and could eventually be applied in other contexts, such as controlling a TV.The hosts then discuss Trump’s proposed $100K fee on H-1B visa holders, unpacking the back-and-forth around the policy. They explore who it impacts most and why Trump is pushing it. While big tech will likely absorb the cost or pivot to other visas, and hospitals may be exempt, early-stage startups will feel more pressure to hire domestically, since $100K can be a substantial part of their runway.Finally, they decide if Oura would “Get the Check.” With Oura allegedly raising $875 million at an $11B valuation, the hosts analyze the company’s distribution strategy, including partnerships with the DoD, Amex, and women’s health companies.00:00 Intro05:51 Meta Ray-Ban Display at Meta Connect29:34 Trump imposes $100K fee for H-1B visas50:15 Oura raises $875M Series E at $11B
This week on Get the Check, the hosts talk about how 1000 gallons of water flooded Maya’s apartment building and how ChatGPT may have saved her $11k.Anthropic just dropped $1.5 billion to settle with the Authors Guild after getting caught training Claude on pirated books from LibGen. The hosts talk about the details of the case and when copyrighted works fall under “fair use” versus not. They also talk about how this impacts other big lawsuits like NYT vs. OpenAI. Finally they speculate why there wasn’t a bigger reaction to this decision.Next, the hosts talk about the browser wars. You might be getting deja vu from when they recently covered it, but there are some fascinating updates.Google beat a massive antitrust case and gets to keep Chrome, even though they still control 94% of search. The judge agreed they have a monopoly, but with the rise of AI sees the future of search as competitive even with Google owning Chrome. Atlassian just spent $610M cash to buy the Browser Company. The deal is slightly ironic, since the Browser Company was all about trendy AI-forward consumer browsers and now is owned by a (almost?) legacy enterprise software player. The hosts disagree on if Atlassian made the right bet.Finally, they spotlight Exa, which is Google for AI agents. The company just raised $85 million at a $700 million valuation and already has customers like Notion and Cursor. The hosts talk about why AI agents need a different type of search tool than humans. All the hosts agree on if Exa would or wouldn’t “Get the Check,” but you’ll have to tune in to know the answer.If you enjoyed this episode please give us 5 stars on Spotify or a thumbs up on YouTube :)00:00 Intro 02:14 Anthropic pays authors $1.5B 20:21 Google anti-trust case ruling, Google keeps Chrome29:41 Atlassian buys the Browser Company for $610M40:40 Would Exa Get the Check?
The hosts start with Trump’s firing of Federal Reserve Governor Lisa Cook over alleged mortgage fraud. They unpack the legal battle around whether a president can remove Fed officials “for cause,” what that means for the Fed’s independence, and the potential implications for monetary policy, markets, and democracy. They draw historical comparisons, from Nixon and stagflation to the UK’s pre-1997 monetary policy, and note Trump’s push for rate cuts to offset tariff-driven inflation.Next, they dive into Open Evidence, an AI healthcare startup backed by top VCs. They explain how the platform helps physicians synthesize peer-reviewed research, why its bottom-up adoption strategy worked, and how it addresses physician shortages and outdated workflows. They highlight its rapid growth, 40 percent daily physician usage, and medical licensing exam success, agreeing it would “get their check.”Finally, they cover SpaceX’s recent milestones: the successful Starship Flight 10 test where both booster and upper stage landed in splash zones, and the Starlink 17-7 mission deploying satellites with a new “Pez dispenser” system. They discuss rocket reusability, Starlink’s growing subscriber base, and how its profits are funding SpaceX’s Mars ambitions. They also explore the challenges of colonizing Mars, from human health risks to international policy questions.00:00 Intro 02:01 Trump fires Fed Governor Lisa Cook 23:42 Open Evidence 39:22 SpaceX's launch
This week the pod dives into the multi-trillion dollar question: is AI a bubble? The hosts break down the history of the dot-com bubble and compare what’s similar and what’s different this time. They also discuss the widely cited MIT study showing that 95 percent of organizations are getting no returns from generative AI, covering which types of implementations are seeing real value and where it actually makes sense for businesses to invest in AI.The conversation then shifts to the latest meetings about the Russia-Ukraine war. Trump has said the conflict would never have happened under his watch and even said that solving it could be his ticket to heaven. The hosts unpack the reality of Putin’s demands for land and his refusal to meet with Zelensky. Although Putin and Zelensky still haven’t met, Trump’s meeting with Zelensky and other European leaders was not as awkward as last time when JD Vance asked Zelensky if he’s ever said thank you.
Finally, the hosts cover whether Eight Sleep would “Get the Check” after raising a $100 million Series D. The cult-favorite sleep tech promises better rest with its smart mattress cover. Tech bros ever might be needing its calming effect if the AI bubble really does pop. Tune in to hear why the hosts disagree on whether they would write the check.
00:00 Intro
01:27 Sam Altman on the AI bubble, 95% of biz have no GenAI returns (MIT)
34:10 Trump tries to solve Russia-Ukraine
48:58 Eight Sleep raises $100M
This week the pod sat down with Minh Le, one of the researchers behind the viral AI safety research study “Subliminal Learning: Language models transmit behavioral traits via hidden signals in data.”
The paper showed that if you have a teacher model with a love for owls that teaches a student model a series of random numbers, the student will also inherit a love for owls as long as they share the same base model, which means models can inherit misaligned traits from other models even if it’s not observable in training data.
The hosts deep dive into the paper’s methodology and ask about Minh’s strategy when filtering out numbers that might carry unintended associations like 666 or 911 that have an association with evil or danger. Fun fact the original plan was to use a love for eagles but they switched it to owls because there were fewer associations that could create potential noise. They also go over theories about why the teacher’s behavior is transmitted when the data transferred is random and filtered. Spoiler, it probably wasn’t a secret code in the numbers but rather the data distribution triggering emergent behaviors in the student model like a love for owls.
The pod also gets into what the media got wrong about the paper, AI safety, and Minh’s hot take on why he doesn’t buy into p doom (the idea that AI leads to human extinction…). Minh also talks about how he went from being an independent researcher to the prestigious Anthropic Fellowship and now a full time role at Anthropic.
00:00 Minh's career journey
04:36 Deep dive into the subliminal learning study
26:23 Larger discussion about AI safety
You might remember Maya decided her life had no value after the pod read AI 2027 a couple of months ago, which outlines a doomsday AI scenario where the human race goes extinct.After this week’s OpenAI GPT-5 launch, Maya has hope again. Sam Altman says GPT-5 is as smart as any PhD student, but the pod discusses why many people aren’t buying into the hype, especially since OpenAI’s own announcement had incorrect charts that GPT-5 generated. Although the model is still doing well on benchmarks.The hosts also discuss why OpenAI is choosing to go with a unified model for the first time, which means user queries will automatically be routed to a longer or shorter “thinking” model behind the scenes.They also discuss the gpt-oss release, which are open weight, something OpenAI hasn’t done in a while. The pod discusses the difference between open weight and open source, and why DeepSeek isn’t actually open source. Additionally, what an OpenAI open weight model means for the industry.Finally, they discuss if Joby would “Get the Check” after announcing it is acquiring Blade’s passenger business for $125M. Blade is keeping their medical organ transportation business and rebranding to Strata.Please like and subscribe so that one day the pod can afford to fly to the Hamptons via Joby.00:00 Intro01:52 OpenAI announces GPT-521:01 OpenAI releases gpt-oss which is open weight34:50 Joby Aviation acquires Blade’s passenger unit for $125M
The year is 2030 your Neuralink chip calls your Tesla back from an EV charging station and your Optimus robot opens the door for you (chivalry isn’t dead).
This week the pod does a deep dive into Neuralink's technology and how the device is able to decode your brainwaves and send them to a model, which translates them into actions like playing a video game. Neuralink already allows Noland Arbaugh who is paralyzed from the neck down to operate a computer with his mind.
Elon’s long term vision for the company is that Neuralinks become as common as having an iPhone, and help humans stay competitive in a world of AI by increasing their brain’s output.
The hosts discuss the recent Neuralink news that has flown surprisingly under the radar: Neuralink is running UK clinical trials and is also developing a smart bionic eye with UCSB to help blind individuals regain sight via a camera-to-brain signal system.
Finally the pod debriefs everything that has happened in the US economy this last week like Trump firing the bureau of labor statistics commissioner because he thought job numbers were too low. They also talk about the takeaways from a spicy Fed rates meeting where two people dissented with Powell on interest rates for the first time since 1993. Finally Maya gives her hot take on why she thinks the dollar is losing value and what that means for the US economy.
00:00 Intro
01:00 Neuralink
32:00 Fed Meeting, Trump fires BLS commissioner
The pod read Trump’s 90 AI policy recommendations so you don’t have to. This week, they dive into the highlights, including why the government wants an open-source model, how to build a grid and workforce that can support AI innovation, and why the administration wants AI to be built on an American tech stack. The hosts break down why both China and the US want developers building on their models… hint: it’s not about money, it’s about soft power — one of those things international relations professors always bring up that actually turns out to matter.They also talk about the Tea app, which hit #1 on the app store, dethroning ChatGPT. This week, the app experienced a data breach that exposed the identities of 13,000 users. The hosts unpack the drama, including how the breach may have led to the exposure of a secret US military base.Finally, the hosts ask if Figma would Get the Check in light of their upcoming IPO. It’s hard to stay unbiased when the company is how co-hosts Priya and Anika met <3 but they do their best to focus on year-over-year growth and Figma’s impressive user retention.
00:00 Intro
00:56 Trumps' AI Action Plan
38:50 Tea App Hack
48:11 Figma's upcoming IPO
If you read the entire internet multiple times over you would be the smartest person on Earth, so why is it that every frontier model has done that and you still have to read over a simple email ChatGPT writes?
Tune into this week’s episode to hear Maya’s answer: AI models are getting trained too inefficiently and Moonshot AI’s Kimi K2 is an example of how training can be more efficient. Kimi K2 is China’s latest open-source model, which was trained for agentic use cases. It’s doing better than anyone expected, and if you’re getting DeepSeek deja vu so are we.
The episode also dives into the AI browser wars and Priya gives her hot take on who she thinks has a right to win between Perplexity’s Comet, Dia by the Browser Company, and ChatGPT Agent.
Finally, the pod debriefs Lovable’s $200M Series A and the future of GPT wrappers.
00:00 AI browser wars
25:56 State of AI in China
52:18 Lovable raises a $200M Series A
What do Love Island and BILT have in common?
They both just had a major breakup this season.
BILT and Wells Fargo officially called it quits, something our OG podcast listeners might remember we predicted back in our very first episode.
So far, BILT is winning the break up. BILT just raised $250M at a $10.75B valuation. Fintech valuations seem to be back, but Maya (aka our resident Fintech guru) has some doubts about how bright BILT’s future in the mortgage market is.
We unpacked why Grok had an identity crisis and started calling itself "MechaHitler”. We also talked about how Grok 4’s benchmarks are impressive, but it still faces an uphill battle for user adoption.
Plus, we break down the drama behind Google's $2.4B deal with Windsurf and what it means for tech employees.
00:00 Intro
01:21 Grok 4 and "MetchaHitler"
27:35 Google's Windsurf deal and the AI talent wars
41:56 BILT's Series C and end of Wells Fargo partnership
🗳️ How did the city with one of the biggest shifts to the right in the 2024 presidential election just vote for a socialist?
This week the pod breaks down Zohran’s rise from 1% in the polls to winning the Democratic nomination for NYC mayor. They discuss how Zohran was able to create a memorable brand and also dive into his most talked about policies:
🚼 universal childcare
🥶 a rent freeze
🛒 a city-run grocery store
💵 raising the minimum wage to $30
Then we get into Elon’s rideshare launch in Austin that he’s been hard launching since 2019, but it was a soft launch with 10 to 20 cars in a fenced-in area, priced at $4.20. Of course. The hosts discuss:
📷 Tesla’s bet on cameras over lidar
🚗 a Tesla dropping a passenger in the middle of an intersection…🤖 Tesla’s other moonshots like the humanoid robot Optimus
00:00 Intro
00:47 NYC mayoral election and Zohran Mamdani
45:56 Tesla's rideshare launch in Austin
This week we sat down with Emir Karabeg, the co-founder of Sim Studio (YC X25), an open-source AI agent builder used by the Department of Defense and Mercor. We debriefed his YC Demo Day and discussed his insane lore, including:
💥 Building a crypto mining operation at 14 that accidentally triggered an MI6 raid
🧠 Why what he learned at YC was life changing
🚗 How he gave away 10 Teslas with David Dobrik (and one extra by accident)
📞 Leaving in the middle of a UC Berkeley final to take a call from Olaf Carlson-Wee (Coinbase employee #1)
00:00 Intro
01:54 Crypto mining at 14 years old
12:34 Dropping out of Berkeley (for a semester)
22:10 YC Demo Day
30:12 Building Sim Studio
40:54 Founder philosophies
51:29 Reaching AGI
Meta just dropped a casual $14.3 billion on Scale AI for a 49 percent non-voting stake. The hosts are wondering if Zuck is in his desperation era handing out NBA sized contracts to anyone that can rescue Llama. Alexander Wang is now running Meta’s new Superintelligence Lab, and the hosts breaks down what Alex leaving means for Scale. In the short term, employees are getting a massive cash payout, but long term Google, Microsoft, and xAI have already said they’re not renewing contracts with Scale, which opens the door for Scale’s competitors like Snorkel, Surge AI, Mercor, and Turing to step in. Maya floats the idea that Scale may be a product play and not just an acqui-hiring of Alex, if Meta has a bigger plan to use Scale to build continual learning models inside Meta.
Then the episode covers the largest single-day protest in American history. Around 5 million people turned out for the No Kings marches in response to ICE raids and Trump’s executive overreach. Anika shares what it was like on the ground at the SF protest. The hosts also unpack the very public back-and-forth between Gavin Newsom and Trump. They also analyze Trump’s immigration strategy in comparison to Obama’s. Finally, they get into the messy economic and ethical logic (or lack thereof) behind targeting undocumented immigrants who’ve been here for decades.
00:00 Intro
01:18 Meta acquires 49% of Scale AI
37:13 ICE Raids lead to the No Kings protest
The hosts get together at 7am after Anika and Maya both managed to sleep through 6am alarms. This week the trio dives into Nucleus Genomics, the hot IPO market, and Mary Meeker’s AI trend report.
Nucelus Genoomics just launched a new product called Nucleus Embryo. Nucleus Embryo lets IVF parents compare embryos based not only on health markers, but also on traits like height, eye color, and even IQ. The hosts explain the science behind polygenic risk scores (PRS) versus traditional embryo screening for rare diseases. Finally they dive into the many controversies around genetic selection including the accuracy, risks of eugenics, potential for ableism, and the implications of a genetically engineered upper class that gets solidified over multiple generations.
IPOs are so back. Circle, the company behind USDC, had a great IPO, jumping from $31 to $115 a share in a week. The hosts explain how Circle makes its money mostly from interest on the billions in reserves backing the stablecoin, thanks to high Fed rates. The discussion also touches on Circle’s failed SPAC attempt and why crypto might be going mainstream. They also cover Omada Health’s more modest IPO and discuss how digital health IPOs are slowly returning after a long dry spell post the pandemic era. Finally they go over potential IPOs rumored for the rest of the year.
The episode wraps with a breakdown of Mary Meeker’s hyped 2025 AI trend report. They dive into 340 pages of charts, graphs, and the word “unprecedented” (which is used 64 times!). The hosts discuss why ChatGPT adoption has been so much faster than Google. They also discuss how the commodization of AI is going to impact the business model of AI players.
00:00 Intro
03:04 Nucelus Genomics
27:14 IPO market including Circle and Omada Health
40:16 Mary Meeker’s AI trend report