News
Entertainment
Science & Technology
Life
Culture & Art
Hobbies
News
Entertainment
Science & Technology
Culture & Art
Hobbies
Finding the Midnight Snacker : Agentic Marketing by @ttunguz
“Feel Good Fun Mix” tops my recommendations on Spotify. Spotify has created over 6,000 such labels by hand. Amazon and Netflix attributed 35% and 75% of their revenue to their recommendation systems. This is a profound & counterintuitive shift in how we think about marketing. In a recent case study led by Aampe, an AI agent with this type of segmentation sent far fewer messages than traditional systems while achieving better results.
I Use AI 100 Times Per Hour by @ttunguz
How often do you use AI? I tracked my Sunday workday to find out. Between 4:30-9:00 PM (with a dinner break), I monitored every AI interaction while handling emails, analyzing data, & writing. If the average American picks up their mobile phone 144 times per day & we call that addiction, I am using AI about a hundred times per hour. Is AI ten times more valuable than a phone?
Top 10 Posts of 2024 by @ttunguz
These are the top 10 posts of 2025. I like to do a roundup each year as a retrospective of how much the industry has changed through some of these posts from the beginning of the year. Some are outdated already & some presaged accelerating trends. AI Drove the Largest New Bookings of Any New Product : In January, ServiceNow announced that their AI product was the fastest growing product in their history.
Why I Let AI Grade My Writing (Just Like My AP English Teacher) by @ttunguz
Ten years ago, I hired an editor to grade my blog posts as an AP English teacher might. Yesterday, I asked an AI to do the same thing. The verdict? A solid B-. Claude’s Scorecard Grade: B- Strengths: The piece demonstrates a clear understanding of complex technical concepts and attempts to make them accessible to a general audience. The author’s personal connection to deep learning through graduate school studies adds authenticity and historical perspective.
The Year Algorithms Learn to Act by @ttunguz
In the bustling tech campuses of 2024, the age of passive AI – systems that merely respond to our queries – is giving way to something far more profound: the era of AI agents. As we look to 2025, we’re about to discover what happens when algorithms learn to act. At the heart of these emerging agents lies a trinity of learning approaches : supervised learning : like reading a book to a child, humans provide clear guidance to AI labeling cat & dog, sheep & cow.
Theory Two by @ttunguz
Today, we’re announcing our second fund of $450m to support our mission of partnering with early stage software companies that leverage technology discontinuities into go-to-market advantages. This marks the next chapter in our firm’s evolution. Since we launched Theory in 2023, we’ve gathered a wonderful team : Lauren, Spencer, Andy, Rafa, Amber, Arjun, & Kristin. We have partnered with 8 marvelous founding teams, all using data to power the next wave of innovation across the Modern Data Stack, Artificial Intelligence, & Web3.
My Little Library by @ttunguz
I didn’t notice it at first but there in the back corner of my laptop, I’ve been assembling a little library. The library doesn’t contain books, but scraps of text that explain what an AI should do. Visit Anthropic’s home page & you’ll find their collection : Each of us will assemble these little libraries with tracts like : write a blog post in my voice about this topic collect action items to send to the team from team notes & format them in this particular way analyze the P&L for the finance team paying particular attention to gross margin write a performance review using the company’s stated evaluation rubric for my teammates summarize a competitor’s webpage every month to detect changes in positioning As we work in a role, we have assemble these workflows and kept them to ourselves.
75 Cents per Month by @ttunguz
What it cost to have an assistant with you like in the movie Her? The cost of using AI has dropped precipitously, an order of magnitude every year. If the average American picks up their phone 144 times per day & engages with an assistant, each time for four interactions every day of a month, an assistant like Her would cost about 78 cents in inference cost.1 I’m not taking into account any of the additional costs associated with delivering such a product.
Small but Mighty AI by @ttunguz
77% of enterprise AI usage are using models that are small models, less than 13b parameters. Databricks, in their annual State of Data + AI report, published this survey which among other interesting findings indicated that large models, those with 100 billion perimeters or more now represent about 15% of implementations. In August, we asked enterprise buyers What Has Your GPU Done for You Today? They expressed concern with the ROI of using some of the larger models, particularly in production applications.
The Post Election Surge is Unevenly Distributed by @ttunguz
After the election, the public markets have roared, but not equally. The broad software ecosystem has seen a relatively muted change in forward multiples. There’s no statistically significant change in the days after the election compared to the month before. On the other hand, crypto’s top tokens have seen tremendous appreciation. Bitcoin is up 48% ; Solana up 70% ; & SUI up 324% in the few days since the announcenment.
I Talk to Robots While Driving by @ttunguz
Over the weekend, I found myself in an hour-long conversation during my drive with an AI. We jumped from discussing Cooper Flagg’s basketball stats at Duke to comparing Carlo Rovelli and Brian Greene’s competing theories of physics to talking about the history of San Francisco. In a daring feat of economic analysis, I asked it to calculate if the after-tax returns of two ETFs were statistically significant & to compare the energy portfolios using the 13-Fs of a few hedge funds for investment ideas.
Profit Dollars per GPU Dollar by @ttunguz
“AWS’ AI business is a multibillion-dollar revenue run rate business that continues to grow at a triple-digit year-over-year percentage and is growing more than 3x faster at this stage of its evolution as AWS itself grew, and we felt like AWS grew pretty quickly.” “Our AI business is on track to surpass an annual revenue run rate of $10 billion next quarter, which will make it the fastest business in our history to reach this milestone.
The White Collar Revolution by @ttunguz
The major areas of AI innovation automate white-collar work. Reviewing the BLS’ data on employment for white collar work, I aggregated the data to these categories. It’s striking that most of them already have a significant number of AI startups pursuing their ambitions to change workflows. Occupation Employment (in millions) AI Technology Software Developers & IT 2.71 Code completion, generation, refactoring, security analysis Education & Librarians 2.37 Computer adaptive instruction & testing Engineers 1.
Trends in the Post-Modern Data Stack by @ttunguz
On Monday, at TC Disrupt Colin Zima CEO of Omni, Jordan Tigani CEO of Motherduck, Daniel Svnova CEO of Superlinked & Toby Mao CTO of Tobiko Data who are leading the evolution of the Post Modern Data Stack discussed the trends they are seeing. Here are some of the themes & predictions from the group. Customers are excited about new architectures that significantly reduce cost. In the last 10 years, investments in big data have become increasingly expensive & focused on very large data volumes.
My AI Rube Goldberg Machine by @ttunguz
In yesterday’s post, I calculated the profitability of public software companies. To calculate these figures, I built a little Rube Goldberg machine. I didn’t download the data into Excel. Instead, I complexified things by sending the analysis to 4 AIs to see if they would agree. The inspiration : many companies have used Amazon’s Mechanical Turk to crowdsource tasks, & pick a consensus answer across three workers to improve accuracy.
Productivity One Year from Now by @ttunguz
If AI continues on its current trajectory or accelerates, what will change in your business? We’ve been asking leaders of companies & departments this question & the answers aren’t clear. In a few years, consensus agrees that rote work of BDRs & paralegals & software engineering will be somewhat to mostly automated. But determining the timing of that impact is much more difficult : it depends on the accuracy of the AI.
AI Prompts as PRDs : Why Prompts Will Become Important IP Assets by @ttunguz
When I was a novice product manager, I remember hearing that acronym for the first time : PRD. The Product Requirements Document. The PRD contains the output of a conversation between product & engineering - what is to be built within a few leaves of digital paper. Perhaps similar to the conversation one might have with a chatbot about a product feature. 😏 Over the weekend, a Reddit user asked the Meta AI model powering WhatsApp to reveal its system prompt - the instructions Meta engineers provide to an AI as part of every query.
The Premise of a New S-Curve in AI by @ttunguz
Since July, have you noticed how much better your AI model has become? Measuring them is hard to do. All we can do is quantify the vibe : is this one better than that one? Elo is a score that measures how often one model wins against another, as judged by a human. Which model answers the prompt : “Describe the differences in texture between a Pink Lady and a Macoun apple” better?
Why Lifetime Value is Relevant Again in Software by @ttunguz
Budget cuts followed interest rate hikes in 2022. By late 2023, more than a year of financial scrutiny had challenged many publicly traded software companies. However, 2024 has been a tough year again. Net dollar retention for publicly traded software companies fell from 113% on average to 108%, a five percentage point drop.1 80% of publicly traded companies saw their NDR wilt again in 2024. Overall account expansion remains roughly 5 percentage points above inflation.
A Challenge to SaaS Orthodoxy by @ttunguz
Klarna, the Swedish fintech giant, is making waves by churning from industry-standard software like Salesforce and Workday in favor of building its own internal systems with AI. After their success with AI customer support automation which manages 2/3 of their customer inquiries, Klarna is now doubling down on this strategy. Klarna is betting AI-enabled software is the future of internal tools. The corollary : the overall cost of building internal software with AI is lower than buying off-the-shelf solutions.
How M&A Fosters Innovation by @ttunguz
Recently, Thomas Laffont of Coatue highlighted a non-obvious impact of the M&A slowdown after this preamble : “Ironically and I think somewhat perversely one of the byproducts of constraining big companies from buying small companies is it hurts small companies. This point is clear & straightforward. Less demand for an asset decreases price. 2023 & 2024 are two of the three lowest in the last 12 years in both activity & value across the ten largest strategic software buyers.
Fulfilling Crypto's Original Promise by @ttunguz
Visa announced their plans to launch a stablecoin with BBVA today. Throughout the quietness in crypto created by the wake of the FTX collapse & the AI boom, stablecoins have become a large & very fast growing part of web3. Users clock 3m stablecoin transactions per day at an approximate average value of $5,000. But, the use of stablecoins extends far beyond crypto trading. Stablecoins own more US Treasuries than South Korea & Germany & have become an important buyer of US goverment debt.
Where is the Budget for AI Coming From? by @ttunguz
Morgan Stanley surveyed a group of CIOS to understand the sources of AI budget. If hyperscalers deploy $100b in capex on AI this year & legions of software vendors hawk AI solutions, the money to justify it must come from somewhere. 41% of CIOs said net new spend. 35% said existing software spend. Only 6% said it would come from professional services. This last point raises the question about labor replacement : will AI siphon funds from labor budgets?
Would You Listen to AI Generated Podcasts? by @ttunguz
Recently, Google launched the ability to generate a podcast conversation from an article. I was curious if it worked & I tried it out on three recent posts. The output could fit into any radio show or podcast I’ve listened to. It’s not perfect: there are no intros or outros. But the “hosts” chuckle, ask questions, interplay in human-like ways. The AI even references images in the post. The Ghost in the Machine : Invisible Ads in Generative Search In this episode, the AI injects a related concept that I don’t mention in the post : native ads.
Interwoven with Initia by @ttunguz
Like in web2, building an app on a blockchain requires several layers & components. Picking the best of each can be a challenge for a developer. Once the stack is fixed, a developer might want to choose the best database characteristics for an app : lower latency for a game, greater security for a financial exchange. Last, interoperability with other systems is a critical way of enticing users to use the apps.
AI Advantage for Startups : Changing the Workflow through Services by @ttunguz
Suppose you’re a startup in a competitive market with a large incumbent who owns the system of record - the software that runs the sales team or the support team or the marketing team. How do you win? In the last decade, startups have chosen to identify a feature or workflow to improve & leverage that wedge into an advantage. Many have reached great levels of success, but few have overturned the incumbent.
Will We See a Better Exit Market Next Year? by @ttunguz
The Fed cut rates by 50 basis points this week. A mantra circulating in Silicon Valley has echoed that the tepid exit markets will revive as a result of a laxer monetary policy. The last ten years’ data suggest the relationship is real & non-linear.1 When the Fed cuts rates - negative changes in the Fed Funds Rate (FFR) - US venture backed software exit activity increases by between 10% and 65%.
Writing Software for Robots by @ttunguz
In a few years, most feature flags & linting tools & other developer tools will be implemented predominantly by robots. Already, 50% of code at Microsoft & other large internet companies is written by AI. This idea expands far beyond developer tools. Robots will manage sales development, paralegal work, medical intake, & many other tasks. The tools of the last 15 years have been built to drive productivity. But in the future, these tools will drive robotic productivity rather than humans.
2024 Theory Ventures Go-to-Market Survey: Optimism Rises Amid Changing Market Dynamics by @ttunguz
Last week at SaaStr, we unveiled the results of the Theory Ventures 2024 Go-to-Market Survey. This annual survey shines some light into the state of SaaS sales & marketing, offering a glimpse into how founders & companies are navigating the current business landscape. Here’s a breakdown of the key findings: Increased Optimism & Steady Fundraising Expectations Despite the challenging economic environment, founders are more optimistic than 2022. The average outlook score has risen from 6.
The New Software GTM Playbook by @ttunguz
The first version of the SaaS GTM playbook was written twenty years ago. Salesforce championed it. Mark Roberge published The Sales Acceleration Formula about Hubspot’s journey. Over the next two decades, we analyzed, quantified, instrumented, & optimized many aspects of the SaaS GTM. Account executive to SDR ratios, sales cycle lengths, conversion rates, customer acquisition costs, customer lifetime values, net dollar retention. The new software GTM playbook has yet to be written.
The Rarity Shibboleth by @ttunguz
Large language models are wonderful at ingesting large amounts of content & summarizing. Uploading an academic paper and I can pester it with an infinite list of questions & it will respond with equally infinite patience. In comparing the two most recent Microsoft earnings calls, Claude highlighted: faster than expected Azure growth (29% vs 27%) AI contributing 8% of Azure revenue up from 7% higher CapEx spending & greater capacity constraints for data centers much better commercial bookings growth : 17% vs 13% Excellent analysis.
The Challenge of the AI Demo by @ttunguz
The AI Demo isn’t easy. Many of the major AI companies have demoed their AI systems, first starting with pre-recorded, & now pushing into live demos. They don’t always work. Multiply Murphy’s Law by a non-deterministic system & it’s not unreasonable to expect AI demos to nearly always hiccup. Demo disruptions aren’t disaster. These systems are early & changing rapidly. They might suggest the system requires work & tuning, not a fundamental challenge.
Which Design Era Are We In? by @ttunguz
When the internet became popular in the 1990s, websites were static & adopted the design language of their creators : blue links & images. Then, we added interactivity with Javascript. In 2005, Gmail claimed the title of most sophisticated app. Soon there after, Google Maps enabled interactivity in the browser for 2D maps. Next, the first mobile apps were launched. FourSquare introduced real-time location, which persists as a feature in Snapchat & Uber
The Dislocation Between Public & Private Web3 Markets by @ttunguz
With the rise of AI, web3 has become a quieter part of the venture ecosystem. I was curious about the current state of affairs. The launch of the Bitcoin ETF has amassed roughly $58b. Meanwhile, the Ethereum ETFs has about $500m, two orders of magnitude smaller. The some major Web3 tokens have done pretty well in 2024 ; others have underperformed the Nasdaq (+19% YTD). Token / Index YTD Performance, % BTC 37 SOL 36 QQQ 19 DOGE 11 ETH 9 Series A valuations have increased through 2022 and since have seen a much higher variance, with both meaningfully higher & lower prices than the trend line.
Higher Levels of Abstraction by @ttunguz
Over the weekend, Andrej Karpathy shared this tweet & it inspired me to conduct the 2024 GTM Survey Analysis this way. I use a language called R to analyze data because of its ability to generate pretty charts & the depth of its statistical analysis tools. Within 90 minutes I had found 9 key data points in the data that were statistically significant & runs more than 50 analyses. In the past this kind of work would have taken me 30 to 40 hours.
What Has Your GPU Done For You Today? by @ttunguz
A year ago, enterprises balked at the prospect of deploying AI. The dominant blocker : security. By using AI, would my company lose its data as employees passed sensitive queries to large language models? Today, buyers are more familiar & have security options : deploying AI on virtual private cloud architectures, tools to delete data from cloud AI vendors, dedicated security tools for AI, & a panoply of open source alternatives.
AI Public & Private Premia for AI Companies by @ttunguz
We’ve been tracking the performance of publicly traded AI companies since the beginning of the year. Publicly traded companies with AI products or strategies trade at about twice the forward multiple of non-AI peers.1 Within the private markets, the same is true within the Series A. GenAI startup companies raise at about 1.5-2x the post-money valuations of all software companies.2 These businesses represent about 30% of Series As in 2024.
Things that Used to be Impossible, but are Now Really Hard by @ttunguz
Kevin Scott, CTO of Microsoft, framed the opportunity in AI this way : work on problems that used to be impossible, but are now really hard. My immediate reaction : what technical problems are now possible but still hard? What can computers achieve by themselves that two or three years ago would be intractable? But, this point isn’t purely about technical innovation. What business problems are newly solvable? Nearly 10 years ago, I wrote a post about the minimum viable average contract value to justify a sales team.
The PostModern Data Stack by @ttunguz
If the Modern Data Stack is dead, what replaces it? The Postmodern Data Stack. Post-modernism is a broad movement that developed in the mid- to late 20th century across philosophy, the arts, architecture, and criticism which marked a departure from a belief in grand narratives & an acceptance of pluralism. This idea is taking hold in data. Two juggernauts, Snowflake & Databricks, are challenging the dominance of Google, Microsoft, & Amazon at the biggest scales of data.
Snowflake's 4 Areas of Investment in AI by @ttunguz
In a recent Office Hours, Baris Gultekin, Head of AI at Snowflake & a friend from Google, shared his insights into Snowflake’s approach to AI. It was a great conversation to understand how one of the leading data companies is pushing AI forward. The video for the session is here & the audio is here. There are four major pushes for Snowflake within AI. Bringing AI to Data with Cortex. Cortex is a suite of AI building blocks that enable customers to leverage large language models (LLMs) & build applications.