the AI tools every founder and operator should have on their radar right now coding agents - amp (agent threads linked to git commits, useful as shared team context) - antigravity (another way to work with Claude and Gemini) - opencode (open source, multiple models, runs in your terminal) - claude code (subagents, skills, project memory, MCP) - codex (a second opinion from a different model family, great for code review) - cursor (multiple models in one editor) running agents across the company - QM (YC's open-source internal agent, lives in Slack) - multica (AI agents as teammates, with actual task management) - paperclip (a meta-harness for coordinating agents across different tools) - herdr (manage agents running across multiple machines from one place) - buzz (Jack Dorsey's workspace for people, agents, and code) agents that use your computer - claude cowork (desktop tasks and workflows) - dots (OpenAI's always-on agents inside ChatGPT) - muse (Meta's agent for computer tasks) - gemini spark (Google's background agent for Ultra subscribers) - grok bot (one of the more interesting interfaces in this category) tools for actual business workflows - fin / decagon / sierra (customer support) - ramp (finance and spend management with AI built in) - rogo / hebbia / alphasense (research and analysis) - reducto (document parsing, including messy real-world files) - clay (GTM research and enrichment) - granola (meeting notes that are actually useful afterward) - n8n (connecting tools and automating workflows) running AI on your own hardware - ollama + LM studio (running local models with more control over your data) - deepseek / qwen / kimi / GLM (open-weight models for high-volume workloads) - hermes (Nous Research's agent with persistent memory and local deployment options) - moshi (accessing agents from your phone) what am I missing?

most founders know they need to target the right VCs when fundraising but there's something else worth paying attention to: what kind of investor are you actually talking to? besides the obvious differences between private equity, hedge funds, VCs, and other investors, every VC has their own investment thesis, preferred stages, sectors, and geographies but even within venture, investors approach deals very differently I'd break them down into 7 types worth understanding: 1. lead investor. usually writes a larger check, negotiates the terms, and takes a meaningful ownership stake. often does extensive due diligence and may want a board seat or an active role in the company's development 2. follower investor. prefers joining rounds led by someone else. they still do their own evaluation, but often rely on the lead to set the terms and coordinate the round 3. angel investor. invests personal money, often in founders they know or industries they understand well. some are professional angels, while others invest alongside their main job or business 4. institutional VC. invests capital raised from LPs through a professionally managed fund. usually has a more structured investment process, with partner discussions, investment committees, and specific ownership or return targets 5. strategic investor. brings something beyond capital that can directly help your business grow, whether that's distribution, industry expertise, partnerships, or access to customers 6. family office. manages the wealth of one or more wealthy families. some have dedicated venture teams, others invest directly through family members or external advisers. their decision-making processes can vary a lot 7. syndicate. a group of investors pooling money to invest in a deal, usually through an SPV. in my experience, they can be easier to raise from, especially if you're looking for smaller checks and a faster process. but it really depends on who's running the syndicate one important thing: these categories overlap an institutional VC can lead or follow a round. a family office can invest strategically. an angel can organize a syndicate the point is to understand how each investor makes decisions, what they expect from a deal, and what role they can realistically play in your round that can make a huge difference in how you approach them and structure your fundraising

if you're integrating AI into an existing company, learn from the people actually doing it software - @sarahmsachs (head of AI at notion) - @fergal_reid (chief AI officer intercom, building fin) - @fnthawar (VP head of eng shopify, AI-first engineering) - @dharmesh (co-founder/CTO hubspot, publishes simple.ai) - @karrisaarinen (ceo linear, automation loops inside the product org) - @levie (ceo box, enterprise AI adoption) - @cramforce (CTO vercel) - @briansolis (head of global innovation servicenow) - @robdthomas (SVP ibm) fintech - @emilygsands (head of data & AI at stripe) - @rahulgs (CTO ramp, agentic systems in production) - @geoffintech (CPO ramp, wrote the company-wide AI adoption playbook) - @daschreiber (ceo lemonade, AI in insurance) data & defense - @ssankar (CTO palantir) - @2112Power (founder hadrian, automated factories for defense & space) mobility - @praveenTweets (SVP CTO uber, 70%+ of PRs now from agents) - @aelluswamy (leading AI at tesla) - @dmitri_dolgov (co-ceo waymo, physical AI at scale) - @qasar (ceo applied intuition) who else should i add?

some interesting numbers on what's actually happening with AI adoption in companies > only 37% report any impact on EBIT from AI, unchanged from last year. and just 6% qualify as high performers, attributing at least 5% of EBIT to AI > last year, 32% expected AI to reduce their workforce. this year, only 14% say it actually happened > AI spending has almost doubled in a year, from 1.7% to 3.3% of revenue. over 80% of that spending now happens outside IT budgets > 42% expect to give AI agents actual decision-making authority by 2030. only 5% have the governance in place to do it safely > 32% have already decided against buying at least one software product because they could build it themselves with coding agents > around 20% say AI operating costs, including token usage, are already limiting adoption

the best fundraising strategy is to ask investors for advice tell them you've been working on something and have a hypothesis you'd love their thoughts on. they've seen hundreds of startups, so you want to know if you're thinking about the market the right way use the conversation to show how well you understand the problem, the market, and what you're building if you're onto something genuinely interesting, they'll probably ask if you're raising anyway. that's literally their job, finding great companies to invest in and if they're not interested, you still win. you'll hear their concerns, get a chance to address those concerns, and be better prepared when you talk to other VCs

facts that were true in 2020 > the average founder of a top 0.1% startup was 45 > 50-year-old founders were 2x more likely to succeed than 30-year-olds > founders in their early 20s had the lowest chances of a successful exit > even in tech, the average founder was around 40 > 3+ years of industry experience doubled your odds of building a top startup > founders who successfully exited were 46.7 on average facts that are true in 2026 > average AI unicorn founder age dropped from 40 in 2020 to 29 in 2024 > AI startups reach unicorn status in 4.7 years vs ~7 years historically > outside AI, unicorn founders are still getting older > health and biotech unicorn founders are older than those in other sectors > 40% of unicorn founders are repeat entrepreneurs > 25% of unicorn founders are immigrants

don't have ambitious people around you? build your circle online here are some of the best young builders on X to learn from and connect with - @gajesh (18, engineer eigenlabs, building darkbloomai) - @richardzphotoz (19, media & brand a16z) - @zach_yadegari (19, sold cal ai, building yourpersona) - @DeRonin_ (20, ceo arcane, angel investor) - @jia_seed (22, co-founder jam, ex-big tech) - @im_roy_lee (ceo cluely, raised $15m from a16z) - @sandylikesfrogs (Stanford, experiments with AI agents running businesses) - @0xSigil (Thiel Fellow, founder underdogai, backed by a16z) - @ansonyuu (hardware normalfactoryco, AI + manufacturing) - @mathurahravi (design engineer netflix, runs rabbitholeathon) - @AnnieLiao_2000 (founder solaris_ai, grew buildclub to 70k+) - @_rajanagarwal (AI research openai, ex-Amazon AGI) - @aditabrm (built r-1, reducto's document parsing model) - @schisofrenia (co-host MTSlive, AI interviews) - @hannahsgao (math/CS stanford, ex-openai) who else should i add?

read the summary before bookmarking another article you'll probably never open again learning from other people's experiences in your 20s could save you years of mistakes and wasted time

the most active crypto investors who are early on Cosign: cosign.co/lists/paulklayvc/w… lmk who I'm missing!

the exact same principles that apply to poker also apply to venture capital 1. the right decision can still lose going all-in with pocket aces is usually the right move, even if your opponent ends up hitting the card they need. investing in a great company can also be the right decision, even if it eventually gets written off you always need to think long-term. if you're consistently making bets with positive expected value, you should come out ahead over enough repetitions but you can still lose money in the short term, and that doesn't make you a bad investor 2. trust your intuition only when you've seen the same situation a hundred times intuition should work like a trained neural network. with enough training data, it performs well in familiar situations. but when it encounters something completely new, it's much more likely to get things wrong that's why decisions in unfamiliar territory deserve more time and careful thinking 3. discipline starts with your deal funnel we see dozens, sometimes hundreds, of pitches every day, but we've invested in just over 28 companies you need real conviction before making an investment. there's no reason to deploy capital just for the sake of deploying it at the same time, you can't afford to move slowly when a great deal comes along the trick is knowing how to wait patiently and act fast when it matters 4. remember what founders promised you one of the most useful habits I've developed is writing down the promises and milestones founders give you, then comparing them with what actually happens this works for companies you've invested in, companies you might invest in later, or even just to test your own judgment and with AI, tracking all of this has become incredibly easy if the actual numbers are even remotely close to what the founder projected, that's already a great sign even better if they managed to pivot and achieve much better results in my experience, 95% of the time, the results look completely different 5. know when you're going to fold before the hand starts you should have a clear plan for when to walk away before you even play the hand in venture, this becomes especially important when a portfolio company asks for a bridge round I think the best approach is to evaluate the opportunity as if you were seeing the company for the first time would you invest in this company today, knowing everything you know now? if the answer is no, you probably shouldn't put more money in the fact that you've already invested isn't a good enough reason to invest again 6. be honest with yourself you're going to make mistakes in both poker and investing. the important part is learning from them and constantly improving your judgment one of the GPs at Begin Capital always says something I really like: after your first 100 deals, you love everything. after 2,000, you think everything is terrible. only after that does your radar finally become calibrated, and you start making good investment decisions and there are some great examples of people who went from professional poker to successful careers in VC: @hosseeb, @pratyushbuddiga, @seanlippel, @gametheorizing, @JeffGrossPoker, @phil_hellmuth, @zachcoelius

top 6 mistakes founders make when dealing with VCs (all based on real cases) 1. conflicts of interest even if VCs tell you it doesn't matter and they'll support all their portfolio companies equally, conflicts of interest can get complicated very quickly just look at the public dispute involving Factory AI's @matanSF and @vkhosla investors often have access to your roadmap, customer data, financials, and internal discussions. when the same investor backs your closest competitor, it's reasonable to question how that information is handled before accepting an investment, find out which other companies the fund has backed in your space and whether they plan to invest in direct competitors if there's an overlap, make sure the information-sharing boundaries are clear and reflected in your agreements, especially if the investor wants a board or observer seat 2. investor rights you don't fully understand I've seen founders spend weeks negotiating valuations while paying surprisingly little attention to the investor rights agreement the dispute involving @isaiah_p_taylor and @mashadrokova is a good example of how complicated pro-rata rights and waivers can become the parties disagree about what happened, but it raises an important issue: some rights can be temporarily waived and later restored, while others can be amended with approval from a majority of investors, depending on the agreement that means an investor's rights can sometimes change without their individual consent it's worth going through these scenarios with your lawyer before closing. understand which rights can be amended, who has the authority to do it, and how those changes could affect future rounds 3. giving third parties access to your cap table your cap table contains some of the most sensitive information about your company, including ownership, investment history, and shareholder details @karrisaarinen discovered that @cartainc's secondaries team had contacted Linear shareholders about buying their shares without the company's authorization Carta said an employee had violated internal procedures and subsequently exited the secondaries business @ItzSuds later raised a related concern about how much control companies actually have over their data on Carta both are worth keeping in mind when choosing any platform that handles sensitive company information I'd pay particular attention to how the company uses shareholder data, which teams have access to it, whether that information can be used for other commercial purposes, and how easily you can export it if you decide to leave 4. losing control of your own company owning the largest individual stake in a company doesn't necessarily mean you control it Benchmark was an early investor in Uber, and @bgurley sat on its board for years. after he stepped off, Benchmark's relationship with @travisk escalated into a legal dispute over board seats and governance whatever your opinion of that conflict, it's a good example of how much influence investors can have through board structures the important thing to understand is who can appoint and remove directors, what decisions require board approval, and which decisions can be made without the founder's support also consider how that structure might change after another two or three funding rounds a board arrangement that seems perfectly reasonable at seed can become a serious problem later 5. liquidation preferences and who actually gets paid in an exit @naval co-founded Epinions, which raised $45M from VCs. when DealTime acquired the company, more than 6 million common shares were valued at zero, while preferred investors received equity in the combined company Shopping com, the combined company, went public in 2004 and was worth around $750M by the end of its first day of trading. Naval and other former founders sued, alleging they'd been misled about Epinions' real value the lesson is pretty simple: your ownership percentage means very little if you don't understand who gets paid first and how your shares are valued in an exit before raising, model what happens if your company sells below its last valuation, especially after multiple funding rounds with different liquidation preferences sometimes a lower valuation with cleaner terms is a much better deal 6. terms that give investors all the upside and leave founders with the risk @JosephJacks_ recently shared allegations about a fund offering terms that included a 5x return requirement, full anti-dilution protection plus additional equity, and the right to remove founders if the 5x wasn't achieved within three years I haven't independently verified those terms, but the combination illustrates something founders should be very careful about investors already receive certain protections through preferred shares and investment agreements when those protections become excessive, you can end up in a situation where the investor has significantly reduced their downside while retaining most of the upside and if the agreement also gives them the ability to remove you from the company, the consequences can extend far beyond dilution worth understanding how these provisions work together, rather than evaluating each one separately one more thing I'd strongly recommend is doing reference calls with founders whose companies didn't work out it's relatively easy to find founders who speak highly of their investors when everything is going well the more interesting questions are about what happens when a company misses targets, struggles to raise its next round, or has to shut down how did the investor behave? did they continue supporting the founder? how did they handle difficult decisions? those conversations can tell you much more about the people you're about to spend the next 5-10 years working with

the whole idea behind late-stage investing is that it's supposed to be less risky but what's interesting is that this is becoming less and less true in today's market businesses and business models used to be much more stable if a company had $5M in revenue and 40% margins, you could reasonably expect those numbers to stay relatively stable. they weren't going to change dramatically in a matter of days now that can happen all the time margins can change overnight, customers can quickly switch to a better solution, and an entire business model can become much less attractive in a very short period of time and there's another thing I think a lot of people are overlooking we're now entering a world of trillion-dollar companies but the number of investors who can actually participate in rounds at those valuations is much smaller think about it if you invest at seed and want to exit at Series B, there are plenty of funds that can participate in that round and potentially buy your stake but if you invest at a valuation of several hundred billion dollars and the company eventually reaches $1T, who are you going to sell to? there are very few investors capable of writing checks that large at that point, an IPO might be your only realistic exit opportunity, and the company might have plenty of reasons to stay private so you could end up taking on significant risk at a late stage, while also having far fewer options to exit which makes the traditional idea that later-stage investing is automatically safer a lot less obvious than it used to be

one of the best things you can do as a VC is give founders honest feedback when you pass on a deal here’s why: 1. don't burn bridges if you pass on a company and it ends up taking off, you might want to get into the next round but why would the founder let you in if you couldn't even explain why you passed the first time? 2. good feedback is valuable on its own good founders appreciate thoughtful feedback, especially when it comes from someone who actually understands their market sharing your real reasons for passing shows that you've done your homework, have your own opinion, and aren't just following the crowd and even without investing, you've already given the founder something useful so there's really no point in making up some polite excuse to avoid an uncomfortable conversation just tell them what you actually think

thanks for all the interest! will get back to everyone, just need some time to go through all the messages feel free to reach me at [email protected] also, if you’re in any deal flow chats, would love to join

looking for founders building in deep tech, especially physical AI, AI infrastructure, and AI rollups checks from $500k to $8M drop your thesis, traction, and moat below

one of the highest ROI use cases for personal AI agents rn: finding deals and opportunities that normally take hours of searching, constant checking, and a bit of luck the best ones usually disappear fast now you can have an agent checking for you 24/7 for example: - good seat on a plane, train, bus, whatever. keep checking which seats are still open and grab a better one when it becomes available - reservations at hard-to-book restaurants. instead of checking every day, let an agent watch for cancellations and grab a table when one opens up - good listings for houses, cars, bikes, art, and so on. the best deals can disappear incredibly fast. an agent can watch the market 24/7, filter out the garbage, and send you the few actually worth looking at - closed events. even if you meet all the requirements, spots are limited and applying early can matter. an agent can track new events and get the application in as soon as they open - new collection releases. some are lotteries, but plenty still reward being early. an agent can track launches and make sure you don’t miss them - limited inventory and restocks. sneakers, watches, electronics, collectibles, or just something sold out in your size. let an agent keep checking and grab it when it’s back instead of refreshing the page all day - tickets for something really hyped. concerts, theater, fights, whatever. use an agent for the initial drop or let it keep checking until a good seat comes back from a cancellation - visa appointment slots. not sure how well it’ll compete with professional bots, but it’s definitely better than refreshing the page yourself all day there are probably hundreds of use cases like this, especially once you go into specific niches basically, anything where the opportunity is limited and speed matters use it while the rules still allow it because I doubt they’ll stay the same for long