Perspectives

Autonomy Comes to the Built World

February 5, 2026

Innovation digital world compounded decades Software gets faster AI systems improve exponential pace physical infrastructure tells different story Bridges age housing supply lags demand struggle build factories data centers public works economy requires reason structural limited built safely within given period time severe labor shortages trades required combination high safety requirements persistent scarcity autonomy shines Waymo proved roads Now shift reaching construction Bedrock Robotics leading Today announced 270M financing Emergence thrilled partner team quest Moment Made Click getting know Bedrock's founders Boris Sofman Kevin Peterson something unexpected happened job site autonomous excavator trained scoop dirt dump trucks point empty truck arrived wanted see system Instead idling began rearranging organizing make loading next showed programmed taught scale building became clear remote control extra steps machines work Problem Meets Solution U.S industry short roughly 349 000 workers expects 450 year gap cyclical already slowing starts delayed projects stalled center timelines stretched Meanwhile productivity barely improved remain constrained human availability fixed shifts equipment evolved far slowly environments operates close incremental need new operating model often framed robotaxis heavy ways favorable set conditions Structured boundaries Repeatable tasks measurable outputs Slower-moving workspaces Immediate demonstrable ROI founding including leaders understood deeply applied lessons larger slower central automating easier car navigating city streets payoff Across early deployments value appearing contractors care higher utilization extended hours fewer delays Safer reduced operator fatigue incidents better awareness predictable tighter consistent output planning engaging broader positioning neutral layer existing fleets Caterpillar's toward deployment technology creates models standalone product becomes part done day delivers throughput schedule certainty simultaneously stops experimental essential AI-Enabled Service lens shapes invest services selling outcomes yards moved completed inseparable enables enormous meaningful questions around capture beginning worked closely pricing packaging drawing experience helping companies price delivered rather underlying improves predictability opportunity aligns incentives scales feature business Possible long-term impact excavation happens operate longer consistently coordination teams changes staffed planned measured ultimately believe potential become default platform reshaping cities humans robots directed coordinated ballet Congratulations Laurent entire move

Adaptable AI for an Adapting World

February 4, 2026

last decade progress AI followed simple rule scale Bigger models data compute something work answer usually make larger approach delivered remarkable breakthroughs created systems rigid costly difficult use real world Training frontier become increasingly expensive roughly 2.4x per year since 2016 future runs projected exceed 1 billion 2027 founders met pitched new labs scale-at-all-costs approaches predecessors Sara Hooker Sudip Roy quickly realized build different tune Adaption founded around firm conviction alone longer enough reached insight years front lines production Google Cohere led cutting-edge research building actually run engineered infrastructure delivers reliable scalable performance saw firsthand strain workloads experience revealed truth works practice stay frozen must adapt continuously learn Builders Limits global leader whose spans full stack pretraining proven track record diverse environments Brain advancing across hundreds languages perform reliably real-world conditions Recognized influential voices field combines technical mastery deep commitment creating responsible inclusive serving advisory councils shaping brings rare systems-first foundation builds intelligence powerful fast capable thriving worked core Pathways serve Gemini inference platform efforts working closely hyperscalers enterprise customers knows falters pressure form business understanding matters now ever Enterprises experimenting margins 10 000 organizations putting 11x year-over-year signaling clear shift experimentation operational moves deeper workflows reliability cost control stop nice-to-haves makes Sudip's partnership special complementary backgrounds reimagine together interface design sequential steps afterthoughts co-designed lockstep ground today fixed deployment model trained released expected handle every request fails users instead Prompts grow fragile Costs rise forced reason deeply task invert dynamic kind responds user intent adapts context Rather forcing problem single general-purpose adjust dynamically evolving time means treated equally tasks remain lightweight complex draws reasoning necessary system match workload becomes faster significantly efficient unnecessary choice transforms economics latency complexity size developers running repeatable difference determines whether stays experimental pushed cases industry obsessed brute-force scaling takes path prioritizing efficiency adaptability belief rooted everywhere architecture team assembling Changing Emergence aspire enduring businesses lasting impact stands rarely pattern judgment shows early choose prioritize aim contribute conversations breakthrough contributions given unique lens importantly ambitions limitless thrilled leading Adaption's 50 million seed round investment represents next phase away static toward measured introducing stages starting vision adaptive goal put adaptable hands letting usage shape comes intelligent know strange setting motion

Promoting Our Values

January 22, 2026

Emergence values abstract principles behaviors practice every day shaping invest partner founders build firm long run simple demanding focus drive conviction strive important win big Today proud recognize three leaders whose promotions reflect action Lotti Siniscalco promoted General Kyle Murphy Principal Thomas Kim Head Investment Operations Winning Lotti's promotion reflects always approached investing rigor deep loyalty brings together qualities often treated opposites venture capital pairs intellectual profound competitive fundamentally people-first mindset unapologetic caring deeply invested becoming building looks intensity resilience teams gravitates toward fit single mold show clear slope recognizes ability learn quickly adapt pressure compound time instinct shaped journey first-generation immigrant learned early progress depends willingness think beyond established paths perspective sharpened appreciation operating different constraints reinforced people grow start Known internally work behind develops strong point view market enters pressure-tests assumptions stays closely involved companies evolve finance training informs preparation led several investments particularly fintech insurtech space including Federato recently raised 100M Series Goldman Sachs Casap Method unannounced life outside parent young children deepened leads Parenting importance listening empathy self-awareness influence without force lessons translate directly partners knows push step back trust compounds fiercely loyal shows uncertainty setbacks consistency care years earned long-term taken alongside rare investors want matter weather future uncertain interrogate real issues discover right path forward going cheer team allowing hype blind next great challenge continues already left indelible mark Ross CEO Co-Founder Striving Kyle's role plays true especially moments judgment speed follow-through Since joining grown expanding scope staying relentlessly founder-focused background Plaid experience data infrastructure give practical insight actually takes complex regulated environments turn help navigating critical decisions evaluating payments stacks structuring partnerships recruiting key talent making high-leverage customer introductions question comes goes relevant context follows sets apart competence reliability consistently personal ownership outcomes means founder's beast clearly represents love willing go extension top genuinely nice person energy room Matt Founder Adaptive Underpinning impact team-first lifelong Duke fan alum kind glue-guy rely teammate invisible lifts everyone around Focusing formalizes become essential builds come alone focused approach allows pattern recognition taste development pairing strategy systems ensure tight learning loops helps institutionalize past helped bring discipline Emergence's process Drawing GTM RevOps military veteran strengthened track opportunities surface insights make improved clarity across funnel truly matters title uncommon design created internal processes scale success Congratulations well-earned expression partnership enduring decision

The Invisible Game: What It Really Takes to Build in America as an Immigrant Founder

January 7, 2026

landed U.S twenty years ago alone family friends continent thought hard work enough 16 year old foreign exchange student Italy thick accent suitcase unshakeable belief pushed figure everything wrong Today partner Emergence Capital investing next generation iconic AI companies ambitious founders world immigrants Eric Yuan denied visa eight times getting approved ninth try went build Zoom Vipul Ved Prakash Together.AI William Ross Steenbergen Federato Tony Jamous Oyster name Emcap portfolio came knowing system built anyway drawn ones playing mode see obstacles friction reasons quit something prove investors often best builders take shortcuts assume anything come easy know around complex systems entire lives essay anyone navigating invisible game trying matters underqualified late harder version nobody handed rulebook America Runs Code Silicon Valley loves meritocracy myth great win truth ideas matter execution products rise top merit operate vacuum infrastructure people never arrived apply college Back home pick single path fourteen credit scores Greek life everyone cared sports understand office hours networking events talking accomplishments arrogance worked terrified ask help watched American classmates personal advisory boards tried brute-force judged self-promotion shallow European culture understatement valued building support rigged coded crack remix interesting innovation happens edges outside finally changed finding willing US-plain who'd figured ahead Americans took time translate now Santi Subotovsky similar Argentina network history safety net rent apartment Every VC firm told background right schools clubs invest intensity resilience who've proven break walls said possible separates good understanding relentless effort actively look ways find others miss turn constraints competitive advantages always mindset Real Risk Immigrant Carry worships failure scrappy garage startup pivots three product-market fit founder crashes burns comes stronger badge honor narrative works allowed stay country go founder's company fails maybe join another become advisor investor rebuilding tied job spouse's bureaucratic error funding round takes long close policy shift Washington creates low-grade anxiety fully goes away realize energy drains gone full optimism place felt possibility reality hit disappear almost misstep gave survival dozen different visas status wife employee specific safe jobs traditional finance partly legal predictable wanted risks earlier explore bet uncertain permit depends continuous employment day green card shifted career accelerated make high-conviction bets without hedging existence became citizen arriving seven months pregnant sobbed relief holding little paper flag fog lifted bolder creative decades flooding solid ground beneath feet transformation fundamentally shaped feel impossible big stakes highest someone sees potential proof kind want backs early believes deeply helps need extraordinary else Hedonic adaptation new car bigger sure lose shine permanence freedom wondering denial erase changes Move taught crucial constantly commit weight lift case worth pain recognize choice brings question moving targeting customers answer simply yes head sales leading development dominates adoption Last 109 billion flowed private investment 78 organizations use least business function 55 startups captured nearly half venture deal value enterprise buyers experiment feedback loops fastest scaling actually exists easier start scratch convert traction elsewhere rarely give modest international revenue clean slate sharp go-to-market plan small decision risking ability relationships proximity means regrouping financially mean leaving offers access scale sophisticated second chances countries maximum upside truly cost decide Playbook Wish stages immigration attorney save stress later process moment eligible gap fast slowly secure residency absurd mentors five ten especially navigated met carried crushing mistakes avoidable corners immediately Open soon arrive terrible terms first shape renting loan Create whatever lets sleep night huge savings strong team tight circle call 2 a.m Standing emotional makes swings overnight financial lighter gets eventually room numbers tell story 1st gen population found billion-dollar 65 America's roughly 3.5 likely unicorn born 4 specifically coincidence left behind follow vision already overcome face lifetime adversity fuel problem-solvers opportunities spent environment gotten shifts weekly adding layers confusion Byzantine January 1 2026 nationals 17 complete bans additional 19 partial restrictions including Nigeria populous Africa 5 seeking legally immigrate barred students deportations expressing perspectives social media petitions entrepreneurs trigger page-long interrogatories rejection rates seen Starting moonshot constant review watching added travel ban list courage shipping product midnight filing paperwork Though heart deep respect designed hear Together Lotti Siniscalco invests stage 80 backed 2024

Laying the Foundation for a Billion Robots: Our Investment in Physical Intelligence

November 21, 2025

Demand Robotics Present Urgent Robots long promised answer severe worsening industrial labor shortages real Leaders multiple public manufacturing logistics companies told board-level topic discussion across organizations grapple persistent Today's Fall Short challenge limited scope extremely hard deploy confined simple repetitive highly structured tasks e.g palletizing pick-and-place etc Every new task must painstakingly planned manually programmed arcane PLC code tested extensively on-site combination factors leads sky-high implementation costs often equal greater cost underlying hardware example standard robotic arm heavy industry 150-200K Moreover logic statically hardcoded operate stable environments fail handling deformable materials liquids fabrics cables struggle tackle long-horizon outdoor excluding massive swaths potential use cases Enter Physical Intelligence GPT PI building generalizable foundation model adapt significantly less custom programming manual data collection Early pilots show deployment times shrinking months days transferring quickly robot types profoundly difficult Unlike large language models trained open web internet-scale corpus physical-world ready used training need actual high-fidelity simulations scaling runs far pure vision domains challenges assembled pantheon reinforcement learning enterprise software Karol Hausman Chelsea Finn Sergey Levine among world's cited researchers drawing wealth experience leading labs Google Brain Stanford UC Berkeley Lachy Groom employee Stripe led numerous product growth initiatives issuing checkout international expansion Opportunity Ahead Billion market immense Immediately 30B SI primed disrupt making deployments radically easier tip iceberg PI's enable handle dexterous dynamic make automation finally possible entirely industries huge amounts latent everything textiles electronics healthcare energy Consumer exciting frontier application Emergence laser-focused investing technologies change people work next great transformation excited partner lead charge believe become defining generation helping usher era intelligent adaptable part everyday world

Genspark: The Ultimate System of Context

November 20, 2025

AI infused every software product now Anyone spin wrapper around foundation model ship demo call However gap intelligence impact wide Models good generating answers stop short actually work distinction published System Context thesis companies create lasting enterprise value ones build scaffolding turns generic reasoning specific results Genspark stood immediately already reflects shift Within five months launch company crossed 50 million ARR closed 275 Series 1.25 billion valuation Growth curves happen customers see solving real potential reason behind traction runs deeper speed beginning designed single purpose business productivity simply alternative ChatGPT importantly complete execution encountered Moving Assistance Outcomes tools reduce effort responsibility draft analyze burden guiding correcting stitching everything together falls user never fully understands trying achieve step away process takes different approach company's new Workspace intent describe outcome want delivers finished Presentations analyses spreadsheets research summaries financial briefs complex outputs arrive level autonomy possible surrounding architecture mature Genspark's Mixture Agents coordinates 30 150 in-house 20 premium datasets integrates knowledge workers operate gathers scattered required perform tasks executes workflow hand holding Team platform come first principles alone comes built systems scale founders bring decades experience building search ranking Microsoft Google Meta YouTube Pinterest Co-founder CEO Eric Jing founding member Bing CTO Kay Zhu launched deep neural production COO Wen Sang exited SaaS backed Combinator Khosla Ventures background shows operating discipline confidence investors category Evolution marks important moment shifts peripheral assistant core layer strategic connects operational Individuals teams place captured point action improves time gets done looks practice

All You Need Is Context

November 5, 2025

Love John Lennon wrong least comes AI need context Foundation models incredible write code summarize reason superhuman scale optimized generic use Real work happens inside teams tools systems decisions unique every organization valuable applications next decade build around System integrated architecture owns organizes synthesizes contextual information operationalize AI-powered actions closed-loop learning maximizes value capture defensibility interaction connects everything makes company transforming outputs specific actionable results six main types matter Put simply generate intelligence generates impact Action Strengthens learns output decision feeds back enriching understanding user task environment loop works continuous feedback ensures used smarter precise becomes Moat enterprise chasing question defensible everyone access answer Model surplus creates elevates raw something uniquely useful drives four ways improves quality accuracy fewer hallucinations higher trust enables secure data compounds enriches graph builds proprietary impossible replicate difference companies experimenting generating lasting advantages come deeply workflows VoM essentially attributable measure faster already seeing early versions shift Bolt project44 Mezmo building copilots respond act platforms Zoom connecting customer product workflow pattern clear winners powerful ones complete nothing done needs

Precision Over Promises: Why We’re Backing Vesence, the AI Platform Making Legal Work Bulletproof

October 28, 2025

95 AI deployments fail corporate environments today pattern predictable companies either attempt DIY solutions deploy generic one-size-fits-all platforms overpromise underdeliver different path success depth breadth Obsessing high-value use case drive undeniable results Nowhere focus vital legal profession precision preferred existential Quality Crisis brimming promise text-rich high-stakes field tailor-made large language models tools market stumble simple reason try expense Almost right cut law Hallucinations inconsistencies prompt paralysis turn time-saving innovation liability Lawyers end re-reviewing every clause citation adding hours work new risk project firms tried ensure documents throwing people problem creating armies reviewers whose job scroll Word docs line costly inefficient deeply human-unfriendly Ironically kind actually great Vesence Assurance Era flips model head Instead generating text wrong uses already written Think Grammarly far powerful automatically enforces firm-wide standards flags validates references polishes perfection lives Outlook interfaces prompts learn payoff immediate saved four last night never sending another document without feel intuitive reliable turns control burden competitive advantage Founders Built Persistence Vesence's co-founders Henrik Hansson Ludvig Swanstr embody craftsmanship grit look Emergence founder brilliant technologist Scandinavian eye design entire early platform including complex Microsoft integrations teams need months full engineering squads execute former professional athlete brings discipline speed relentlessness elite competition Together shown obsessive customer sleeping top Swedish firm understand workflows firsthand dedication separates good ones Future Professional-Grade represents next integrate seamlessly quietly make extraordinary team scales expands across thrilled lead 9M seed round help set standard world Welcome family

Expanding Opportunity Through the Emergence Fellows Program

October 23, 2025

believe great ideas come anywhere everyone gets opportunity pursue know career possible without mentors showed achievable want talented energized young people feel access awareness incredible opportunities within entrepreneurship venture capital matter Creating driven professionals inspired launch Emergence Fellows Program new initiative designed empower next generation innovators scholarships mentorship real-world exposure core values developing built team help founders grow extends value outward offering financial support opening doors students otherwise business industry networks proud welcome first class six exceptional UC Berkeley Howard University institutions ranked among top 10 U.S universities advancing social mobility expanding higher education Nearly 500 applied across programs selected receive jumpstart professional development stood curiosity creativity determination use innovation forces change Please join congratulating 2025 2026 Giselle Soto Calderon B.S Administration 27 Tina Le James Nguyen Bioengineering Emanuel Serafin Maciel Molecular Cell Biology Julian Warren Mechanical Engineering Aamori Freeman Computer Science coming months participate meetings sessions build gain familiarity work personal research projects investment theses exploring emerging entrepreneurial born campuses culminate summit share insights recognition goal expand sense long-term begin journeys often reach early small meaningful step toward changing helping skills relationships confidence ambitions excited learn group continue investing potential shape future family

Two Paths for Enterprise AI: Retrofit or Rebuild?

October 1, 2025

next wave AI transformation reshape businesses build maintain core software systems Magic Arrived Developers already experienced moment Tools Cline Claude Code fundamentally changed gets written conversation IDE terminal describe want watch iterates catches mistakes reasons problems alongside fraction knowledge workers people clicking typing navigating ChatGPT question come everyone Enterprise Gap take tangible example Say run sales operations mid-sized company VP told redoing territory assignments quarter First need update lead routing funnels leads right rep CRM new territories configured Finally commission management reps paid correctly test screw lose deal payroll employees needs customized customizations weeks months longer depending complexity Today pointing updates promise vibe business processes Updating Salesforce feel conversational fast intelligent less manual configuration Divergent Paths fork road retrofit yesterday's platforms entirely AI-native Path 1 Incumbents Add Existing develop interfaces users desired changes natural language system proposes modifications workflows custom fields logic interactive experience see proposed accept reject quickly challenge providing legacy architectural sit complex foundations Training understand typical instance requires parsing Apex Flow definitions proprietary XML permission structures countless abstractions built decades 2 alternative radical building ground LLMs excel understanding modifying struggle formats domain-specific languages Instead teaching Salesforce's simply defined TypeScript Python good getting better early signs companies letting configure rules chat interface generates rather endless screens products purpose-built functions Competitive Battleground creates real opportunity startups Incumbent enormous advantages integration ecosystems established communities switching costs border prohibitive burdened makes AI-driven customization genuinely difficult entrants face different equation workflow engines designed specifically go-to-market landscape Emergence Framework believe inevitable five years conversations modify architectures lean models strengths particularly interested Start wedge Target specific process Capture data toward record Use ownership becoming primary subset platform Embrace code-first Enable customers frameworks e.g Provide opinionated typed SDKs string together pre-approved actions UX necessarily underlying Novel UX's display visually represent allowing inspection best opportunities look Trojan horses backed gradually eventually become coding rails massive