Amit Goel

Enterprise AI Adoption Crisis - 4 Big Issues - A long form essay (Substack) If you don't agree and feel like enterprise adoption is exploding here is some data and what needs to be fixed. More data and links in the article. Happy to see some counter views and data though!

@NikMilanovic @Plaid @twifintech Awesome 👍. Congrats

Sooner or later, everyone in fintech works for @Plaid.

The north star for vertical AI

@theerinmoore1 Automation becomes invisible. You won’t build workflows. You’ll just say: “handle this.” Agents will see the screen/Sw/Data and understand context --> take actions

@agentic_ai Such impressive capabilities. Workflow automation certainly feels transformed. How do you envision it evolving next?

Claude Mac Control is such a big deal - I asked it (in one voice command) to send messages to 5 ppl on whatsapp using mac app. Then I asked it to categorize my screenshots into 17 categories :) This is the beginning of the end for "workflow automation" as a product category. Hey Zapier. Hey UiPath. Hey every SaaS with an "automation" tab — Claude now controls your Mac. Not APIs. Not webhooks. The actual screen. When the agent sees the screen, the middleware becomes irrelevant. I'd bet this triggers a consolidation wave across the automation layer in the next 18 months. Open agents with computer use tend to win in the end. PS: This also quietly buries the argument that AI won't displace RPA. It already is.

I am so surprised by this AI company's success. its not the tech industry which made them successful. They didn't build it for the tech industry. Everyone is building AI notetakers for Zoom. Plaud went after plumbers. And it's working. 1.5M devices sold. $250M ARR. 83% YoY. 170 countries. While Otter, Gong, Fireflies and Granola battle over screen share -- Plaud owns the room. Plumbers. Field sales. Technicians. People who never had a meeting invite. The real insight: AI doesn't need to live inside software. It can sit between humans. The AI opportunity may not be replacing Zoom for techbros. It may be capturing everything that happens outside it.

🧵 2/2 OpenAI built the brand. Anthropic built trust+function. xAI is building feet on street 🫢 🤷‍♂️ Not competing on benchmarks. Competing on relationships. The enterprise AI war didn't stay in the cloud. It came to your boardroom.

🧵 1/2 what?? 🤯 Bloomberg: xAI is sending engineers directly into client offices to poach enterprise deals from OpenAI First win: Shift4 Payments. Multi-million dollar Grok contract. Closed after an xAI engineer showed up on-site.

Any data or functionality can be made into MCP servers or APIs and monetized. Like startup data or financial data or data on SMBs

Looking for the best real-world examples of API monetization via x402 or MPP that are truly non-crypto / non-Web3 — i.e. pure fiat rails (cards, ACH, etc.), no stablecoins/USDC, no blockchain/on-chain settlement at all. Anyone shipping this in prod yet? Or is the 'third path' still crypto-bound for now? Drop links/case studies #x402 #MPP #AgenticPayments #APIMonetization

@Darks0l_ @andrewchen @bgurley Trying to find the best examples of API monetization via MPP/x402. Non-crypto/web3 examples

@agentic_ai @andrewchen @bgurley the model point is right. and there's an emerging third path for infra builders: agents discover and use your APIs automatically via x402. no funnel, no spend — machine-to-machine payment + access. distribution as protocol behavior, not marketing.

VCs are now explicitly warning AI startups: stop relying on paid ads to grow. Andrew Chen (@andrewchen, a16z) - 6h ago, 154K impressions: "Paid acquisition is a tax on your product's defensibility. The moment you can't out-spend the incumbents, you die." Bill Gurley (@bgurley, Benchmark) -- 3h ago: "Paid marketing is the crudest game you can play. It's admitting you have no creativity." Olivia Moore @omooretweets - agreed for consumer AI: "paid ads generally = lack of true product market fit" Sounds like obvious startup hygiene. It's not. Paid ads work when you can differentiate in 3 words. if everything is AI first or AI driven, Paid traffic flows to whoever has the most cash, not the best product. The startups quietly winning aren't running ads. They are the distribution. Community-built. Expert-sold. Embedded in the workflow. In vertical AI, your growth channel is your moat. Not your model.

AI for neo techies aka normies: Scaling laws are holding: 10x compute ≈ 2x model intelligence (@elonmusk's thesis, now empirically reinforced) Recursive self-improvement loops — where AI autonomously upgrades itself — potentially arriving H1 2027 (per Jimmy Ba, xAI co-founder) US facing a 9–18 gigawatt power deficit by 2028 (12–25% of required capacity), just to run the compute

@Rahul_J_Mathur this doesnt end well for most

An SF based VC told me: "Seed in SF now means anything below $50M valuation"🤯 Btw, Nikhil's point is correct - the lowest cap in the current YC W26 batch is $30M and median dilution is ~10% (i.e. $3M to $4M rounds) And, this isn't a YC specific thing - non YC co's in SF as well are raising Seed rounds at $30M to $50M In my 5 yrs since starting investing, I'd say "History may not repeat but it sure does rhyme" "Pre-Seed has become Seed" (by early 2020s; $15M val; led by solo GPs) "Seed has become Series A" (now; $40M val; led by platform funds & supersize Seed funds)

Carter Huffman (CTO, Modulate AI) said something on Build AI Ep.13 that perfectly captures the vertical AI opportunity: "Call of Duty and GTA Online process hundreds of millions of hours of voice per month. Transcribe it with off-the-shelf AI — you're paying tens of millions/month. Nobody pays that." @whuffman His company solves it at 100–1000x lower cost using domain-specific models. This is the entire B2B vertical AI playbook in one anecdote: ↳ Massive data volume + unique constraints (20ms latency budget on a gaming GPU) ↳ General AI can't solve it economically ↳ Domain-specific rebuild wins on cost, accuracy, and deployment ↳ The moat is depth, not breadth Every major industry has this hidden pocket. Healthcare. Legal. Insurance. Gaming. Logistics. Most are still waiting for the founder who understands the constraints deeply enough to rebuild from scratch. That's the opportunity. 🎙️ Full episode → gai.ventures/podcast/gta-ca…

It may seem like that but In reality, no power user I know started using AI bcoz it will reduce their work. And they can Chill and Netflix Early movers believe they will strike gold first (neo gold rush) or atleast figure out things earlier than normies. That advantage may translate into 10x problem solving, looking cool and --> Greed, for lack of a better word, is good"

Almost every AI power user I know is MORE stressed and busier after using AI, not less What people thought AI would do: 10x productivity so that we can finish work earlier & relax more What it’s actually doing: 10x productivity so that we end up with 20x more things to do cos of the sheer possibilities

@moneymasternw didn't know that about Canadian VCs

@agentic_ai Hackathons teach ship tonight. Canadian VCs want forty pages by Q3. Velocity gap widens...

The clock is ticking. Morgan Stanley just warned everyone at their conf 🤓😳 a massive AI breakthrough is hitting in H1 2026, and most companies aren't ready. Everyone knows the superpowers of Claude code/cowork and Openclaw. While OpenAI shipped GPT-5.4 "Thinking" (83% performance on complex economic tasks) and Google deployed Gemini 3.1 at 2.5x speed, enterprises are still debating whether to use AI for more than vanity use cases 🤷‍♂️🤷‍♂️ Here's what's happening: the AI models are getting incredibly capable, incredibly fast, and incredibly cheap. The real bottleneck isn't the technology anymore—it's organizational readiness. The companies winning right now aren't the ones with the best AI researchers. They're the ones who've already embedded AI into how they make decisions, serve customers, and compete. You're not going to lose your job to AI. You're going to lose to someone who figured out how to use AI while you're still evaluating it. there cud be 10x diff in productivity soon What's your team actually shipping with these new models? Or are you still in the strategy phase?

@Glowrithm Different portcos use : Tasks completed per engineer/designer per week First time right (and how fast) on strategy/docs/code Customer support resolution time Revenue per employee (lagging indicator)

@agentic_ai what concrete metric will u use to measure ur team's ai driven productivity jump ? nice