StarZero All shows Try the podcast clipper →

Oct 11, 2024 · 30m · saastr

State of the Cloud: The Cloud AI Era with Bessemer Venture Partners

Sameer Dholakia · 11m spoken Timothy Young · 7m spoken Julia Chapin · 4m spoken Kate Jensen · 3m spoken Rami Karabibar · 1m 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

Hosted by Bessemer Venture Partners at SaaStr, this executive panel examines how leaders from Anthropic, Abridge, EvenUp, and Jasper navigate the emerging Cloud AI era. The session delivers strategic insights on multi-model architectures, defensible enterprise moats, organizational change management, and the commercial shift toward work-based pricing models.

How this conversation actually went

Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. How this is scored →

Jason as informed peer 4.5 Guest teaching 1.5 Guest disagreement 0.3 Jason pushing back 0.2
05100:0010:0020:0030:000:52–4:45 · Jason as informed peer 4/10 Opening Remarks and Bessemer's State of the Cloud AI Sameer welcomes the audience, frames the current AI era through historical tech transitions, and introduces Bessemer's billion-dollar AI commitment. The panelists introduce themselves and their respective companies in a warm, collaborative opening.4:45–11:54 · Jason as informed peer 5/10 Agenda Overview: Navigating the Cloud AI Era Sameer prompts the founders on foundation model selection and model fine-tuning across verticals. He connects their multi-model strategies back to the early days of cloud when companies were forced onto a single infrastructure provider.11:54–14:39 · Jason as informed peer 5/10 Scaling Frontier Models and Emergent Capabilities Sameer asks whether future foundation model gains will be asymptotic or step-function leaps. Kate outlines the technical pillars driving progress, while Sameer reinforces the point with an anecdote about Dario Amodei discovering emergent coding abilities.14:39–20:35 · Jason as informed peer 4/10 Establishing Moats and Defensibility in Generative AI Sameer challenges the founders on defensibility and commoditization risk, lightly ribbing Jasper about starting as an API wrapper. The founders outline their respective moats including workflow integrations, distribution, and proprietary data.20:36–23:54 · Jason as informed peer 4/10 Redefining AI Business Models: Selling Work Over Seats Sameer explores how AI forces software companies to move away from per-seat SaaS pricing toward selling completed work. The panelists also discuss expanding into multimodal capabilities such as medical audio and marketing imagery.23:55–29:12 · Jason as informed peer 5/10 Change Management and Services in the AI Era Sameer compares AI adoption to 1990s ERP implementations where change management was paramount. Timothy emphatically reframes AI adoption, warning that treating AI like deterministic SaaS will fail due to its probabilistic nature.0:52–4:45 · Guest teaching 0/10 Opening Remarks and Bessemer's State of the Cloud AI Sameer welcomes the audience, frames the current AI era through historical tech transitions, and introduces Bessemer's billion-dollar AI commitment. The panelists introduce themselves and their respective companies in a warm, collaborative opening.4:45–11:54 · Guest teaching 1/10 Agenda Overview: Navigating the Cloud AI Era Sameer prompts the founders on foundation model selection and model fine-tuning across verticals. He connects their multi-model strategies back to the early days of cloud when companies were forced onto a single infrastructure provider.11:54–14:39 · Guest teaching 2/10 Scaling Frontier Models and Emergent Capabilities Sameer asks whether future foundation model gains will be asymptotic or step-function leaps. Kate outlines the technical pillars driving progress, while Sameer reinforces the point with an anecdote about Dario Amodei discovering emergent coding abilities.14:39–20:35 · Guest teaching 2/10 Establishing Moats and Defensibility in Generative AI Sameer challenges the founders on defensibility and commoditization risk, lightly ribbing Jasper about starting as an API wrapper. The founders outline their respective moats including workflow integrations, distribution, and proprietary data.20:36–23:54 · Guest teaching 1/10 Redefining AI Business Models: Selling Work Over Seats Sameer explores how AI forces software companies to move away from per-seat SaaS pricing toward selling completed work. The panelists also discuss expanding into multimodal capabilities such as medical audio and marketing imagery.23:55–29:12 · Guest teaching 3/10 Change Management and Services in the AI Era Sameer compares AI adoption to 1990s ERP implementations where change management was paramount. Timothy emphatically reframes AI adoption, warning that treating AI like deterministic SaaS will fail due to its probabilistic nature.0:52–4:45 · Guest disagreement 0/10 Opening Remarks and Bessemer's State of the Cloud AI Sameer welcomes the audience, frames the current AI era through historical tech transitions, and introduces Bessemer's billion-dollar AI commitment. The panelists introduce themselves and their respective companies in a warm, collaborative opening.4:45–11:54 · Guest disagreement 0/10 Agenda Overview: Navigating the Cloud AI Era Sameer prompts the founders on foundation model selection and model fine-tuning across verticals. He connects their multi-model strategies back to the early days of cloud when companies were forced onto a single infrastructure provider.11:54–14:39 · Guest disagreement 0/10 Scaling Frontier Models and Emergent Capabilities Sameer asks whether future foundation model gains will be asymptotic or step-function leaps. Kate outlines the technical pillars driving progress, while Sameer reinforces the point with an anecdote about Dario Amodei discovering emergent coding abilities.14:39–20:35 · Guest disagreement 1/10 Establishing Moats and Defensibility in Generative AI Sameer challenges the founders on defensibility and commoditization risk, lightly ribbing Jasper about starting as an API wrapper. The founders outline their respective moats including workflow integrations, distribution, and proprietary data.20:36–23:54 · Guest disagreement 0/10 Redefining AI Business Models: Selling Work Over Seats Sameer explores how AI forces software companies to move away from per-seat SaaS pricing toward selling completed work. The panelists also discuss expanding into multimodal capabilities such as medical audio and marketing imagery.23:55–29:12 · Guest disagreement 1/10 Change Management and Services in the AI Era Sameer compares AI adoption to 1990s ERP implementations where change management was paramount. Timothy emphatically reframes AI adoption, warning that treating AI like deterministic SaaS will fail due to its probabilistic nature.0:52–4:45 · Jason pushing back 0/10 Opening Remarks and Bessemer's State of the Cloud AI Sameer welcomes the audience, frames the current AI era through historical tech transitions, and introduces Bessemer's billion-dollar AI commitment. The panelists introduce themselves and their respective companies in a warm, collaborative opening.4:45–11:54 · Jason pushing back 0/10 Agenda Overview: Navigating the Cloud AI Era Sameer prompts the founders on foundation model selection and model fine-tuning across verticals. He connects their multi-model strategies back to the early days of cloud when companies were forced onto a single infrastructure provider.11:54–14:39 · Jason pushing back 0/10 Scaling Frontier Models and Emergent Capabilities Sameer asks whether future foundation model gains will be asymptotic or step-function leaps. Kate outlines the technical pillars driving progress, while Sameer reinforces the point with an anecdote about Dario Amodei discovering emergent coding abilities.14:39–20:35 · Jason pushing back 1/10 Establishing Moats and Defensibility in Generative AI Sameer challenges the founders on defensibility and commoditization risk, lightly ribbing Jasper about starting as an API wrapper. The founders outline their respective moats including workflow integrations, distribution, and proprietary data.20:36–23:54 · Jason pushing back 0/10 Redefining AI Business Models: Selling Work Over Seats Sameer explores how AI forces software companies to move away from per-seat SaaS pricing toward selling completed work. The panelists also discuss expanding into multimodal capabilities such as medical audio and marketing imagery.23:55–29:12 · Jason pushing back 0/10 Change Management and Services in the AI Era Sameer compares AI adoption to 1990s ERP implementations where change management was paramount. Timothy emphatically reframes AI adoption, warning that treating AI like deterministic SaaS will fail due to its probabilistic nature.

speaking balance: gold is Jason, purple is the guest (3 minute bins)

0:00 · Jason 0% · guest 100%0:00 · Jason 0% · guest 100%3:00 · Jason 0% · guest 100%3:00 · Jason 0% · guest 100%6:00 · Jason 0% · guest 100%6:00 · Jason 0% · guest 100%9:00 · Jason 0% · guest 100%9:00 · Jason 0% · guest 100%12:00 · Jason 0% · guest 100%12:00 · Jason 0% · guest 100%15:00 · Jason 0% · guest 100%15:00 · Jason 0% · guest 100%18:00 · Jason 0% · guest 100%18:00 · Jason 0% · guest 100%21:00 · Jason 0% · guest 100%21:00 · Jason 0% · guest 100%24:00 · Jason 0% · guest 100%24:00 · Jason 0% · guest 100%27:00 · Jason 0% · guest 100%27:00 · Jason 0% · guest 100%30:00 · Jason 0% · guest 100%30:00 · Jason 0% · guest 100%
Sharpest disagreement ▶ 27:35 Timothy Rejects SaaS Adoption Assumptions

Timothy forcefully asserts that anyone expecting enterprise AI adoption to mirror conventional SaaS over the past decade is completely wrong.

 Hardest push from Jason ▶ 15:20 Sameer Probes the Thin Wrapper Vulnerability

Sameer openly questions how applied AI startups can avoid commodity pricing and defensibility traps when foundational models are available to every competitor.

 Biggest teaching moment ▶ 27:55 Timothy Educates on Probabilistic Software

Timothy breaks down why traditional deterministic software mentalities break down with generative AI, requiring hands-on services and prompt change management.

 Jason holds their own ▶ 24:25 Sameer's Historical Analogy on ERP and Systems Integrators

Sameer demonstrates deep domain expertise by contrasting seamless cloud migrations with 1990s on-premise rollouts where services firms earned $3 for every $1 of software sold.
the scores for every segment, with the reasoning behind each
ChapterTopicJason as informed peerGuest teachingGuest disagreementJason pushing backWhy
Opening Remarks and Bessemer's State of the Cloud AI 4000 Sameer welcomes the audience, frames the current AI era through historical tech transitions, and introduces Bessemer's billion-dollar AI commitment. The panelists introduce themselves and their respective companies in a warm, collaborative opening.
Agenda Overview: Navigating the Cloud AI Era 5100 Sameer prompts the founders on foundation model selection and model fine-tuning across verticals. He connects their multi-model strategies back to the early days of cloud when companies were forced onto a single infrastructure provider.
Scaling Frontier Models and Emergent Capabilities 5200 Sameer asks whether future foundation model gains will be asymptotic or step-function leaps. Kate outlines the technical pillars driving progress, while Sameer reinforces the point with an anecdote about Dario Amodei discovering emergent coding abilities.
Establishing Moats and Defensibility in Generative AI 4211 Sameer challenges the founders on defensibility and commoditization risk, lightly ribbing Jasper about starting as an API wrapper. The founders outline their respective moats including workflow integrations, distribution, and proprietary data.
Redefining AI Business Models: Selling Work Over Seats 4100 Sameer explores how AI forces software companies to move away from per-seat SaaS pricing toward selling completed work. The panelists also discuss expanding into multimodal capabilities such as medical audio and marketing imagery.
Change Management and Services in the AI Era 5310 Sameer compares AI adoption to 1990s ERP implementations where change management was paramount. Timothy emphatically reframes AI adoption, warning that treating AI like deterministic SaaS will fail due to its probabilistic nature.

Statements from this episode (13)

Disclosure
Dholakia: Bessemer Dedicates $1B of $4B Dry Powder to AI
“We've dedicated a billion dollars of our four billion in dry powder to go invest in AI companies”
Sameer Dholakia Oct 11, 2024 ▶ 2:24
Assertion Not checkable as stated
EvenUp drafts roughly 1,000 legal demand letters weekly
“We do about a thousand of these letters or 10,000 pages of legal texts that we write every week.”
Rami Karabibar Oct 11, 2024 ▶ 9:08
Assertion Not checkable as stated
Jasper runs about 39 different AI models in production
“In 2019, we started with one model, and today we run about 39 different models in production across the entire customer base, and so we're really LLM agnostic.”
Timothy Young Oct 11, 2024 ▶ 9:47
Assertion Supported
Dholakia: Dario Amodei Revealed Model Coding Abilities Were Emergent, Not Designed
“Dario said on that panel, That they discovered that the models could code. Like it wasn't a design requirement. It wasn't an intentional, the model. It had the capability. It figured out how to do it.”
Sameer Dholakia Oct 11, 2024 ▶ 13:40
Assertion Not checkable as stated
Young: Jasper was the first true application built on an LLM
“In 2019, when the company launched, I think that was a little bit true. We were the first re true application built on top of an LLM we launched before chat GPT.”
Timothy Young Oct 11, 2024 ▶ 16:00
Disclosure
Chapin: Abridge partnered with Epic for native workflow integration
“And what we've been able to do is partner with Epic, one of the largest EMR providers in the US to be able to natively integrate our technologies into existing clinician workflows.”
Julia Chapin Oct 11, 2024 ▶ 18:53
Assertion Supported
Jensen: Intuit uses Anthropic to explain taxes to customers
“Intuit is using Anthropic to help explain what taxes you actually paid and why.”
Kate Jensen Oct 11, 2024 ▶ 19:12
Insight
EvenUp CEO: AI-Generated Demand Letters Build a Proprietary Data Moat
“In our industry in personal injury, the vast majority of these cases are settled privately, where it's very hard to build intuition around what should an injury be worth. So the note that we have that builds up over time is the more of these letters that we wr…”
Rami Karabibar Oct 11, 2024 ▶ 20:06
Disclosure
EvenUp CEO: We Charge Per Legal Letter, Not Per Seat
“Yeah, I'll try to make this quick, but the way when folks think about us, we don't sell software, we sell work. And so if you think about how we're priced, it's based on the legal letters that we write to our customers.”
Rami Karabibar Oct 11, 2024 ▶ 21:06
Assertion Supported
Dholakia: Accenture made $3 for every $1 of SAP software sold
“Accenture made three dollars for every dollar that SAP sold in software because there was so much to do around change management.”
Sameer Dholakia Oct 11, 2024 ▶ 24:58
Insight
Jasper CEO: AI Startups Must Build a Services Strategy Immediately
“If I was building a AI application company, I would, my first piece of advice is build a services strategy immediately. Do not wait.”
Timothy Young Oct 11, 2024 ▶ 27:30
Prediction Not checkable as stated
Dholakia: AI startups without moats face a race to zero pricing
“Defensibility and moats, so critical. If it, if you don't have it will be a bloodbath. You will be in a race to commoditization and pricing going to zero. It will be ugly.”
Sameer Dholakia Oct 11, 2024 ▶ 29:50
Assertion Not checkable as stated
Dholakia: B2B AI applications face severe churn challenges like consumer apps
“We see a lot of consumer apps on the AI side that have a lot of challenge with churn. Same is true for B to B”
Sameer Dholakia Oct 11, 2024 ▶ 30:16
Made with StarZero

Turn any episode into a week of clips.

This entire site, over 400 episodes transcribed, diarized, checked and made playable, runs on the StarZero media pipeline. Drop in your own episode and the podcast clipper finds the moments worth sharing, cuts them, captions them, and reframes them for every feed.

Try the podcast clipper Learn more