Alex de Bold

@graceclarke Well if you're ever in the Toronto area LMK :)

@adebold One of the best responses I've gotten to this. I'm not Canadian quite yet, but certainly soaking up every second I can get up there ;) Great people. Great ideas.

Day 1 💫🔮

@tobi @Copilot @tobi love this!! I just coded a Shopify integration leveraging VertexAi (soon to include Nvidia's digital humans) so this release is the icing on the cake, helping seal the deal for brands.

Agents will become a common way people shop. So today we are releasing 3 tools to make adding commerce to those agents trivial: - Checkout Kit: embed commerce widgets and checkout(!) directly into your agent and chat. This is already being used by Microsoft’s @Copilot. - Shopify Catalog: low latency global product search across millions of merchants - Universal Cart: shop from any store, anywhere, in one cart With these, you can build seamless embedded commerce experiences quickly, starting with all the best products by the best merchants in the world instead of having to sign them all up individually. No need to build a complex new checkout, or dealing with regulatory marketplace rules. Commerce just plugs in and feels seamless in conversations. We believe commerce should be everywhere people are, so we’re making that as easy as possible for the AI age.

This is the same mental model I’ve been using for the new companies I’m building The “how” has increasingly become a commodity. They why, the story telling, and ability to break through the noise is harder than ever.

@chrija Lol - There are a few interesting things going on with 3D and AI. The biggest challenge and opportunity we're tackling is moving this into real-world manufacturing workflows.

Nice easter egg here if you know the “draw the rest of the fucking owl” meme :)

@BrandonMChu Same w/my parents.

Grateful for the life lottery I won. They had it hard so I could have it easy. Thanks mom, dad, and 🇨🇦

@jasonlk Example #1 - the Flashlight app from app store to core feature in IOS.

I don’t mean to sound like a curmudgeon But I feel like a lot of AI products in sales & marketing that have a bit of traction are just going to be features in larger platforms Not that that’s a new thing It’s just that pace of early AI growth may lead us to think otherwise

Wondering how to actually use AI in your day to day? Our team @a16z asked some of AI's biggest names to nominate their favorite products of 2024 - this is a great starting point 👀 Here are their picks! ⬇️ For general AI assistants: - Perplexity (@perplexity_ai) - AI search engine - Anthropic (@anthropicAI) - general assistant, Claude - ChatGPT (@chatgptapp) - for Advanced Voice Mode! For help getting work done: - Granola (@granola) - AI notetaker - Wispr Flow (@WisprAI) - voice dictation - Gamma (@gammaapp) - AI decks, docs, websites - Adobe (@Adobe) - Acrobat AI PDF summarizer - Cubby (@CubbyApp) - collaborative workspace - Every Inc. (@every)- Spiral writing assistant + Cora email product - Lindy (@getlindy) - AI agents to automate workflows To build an audience: - Delphi (@withdelphi) - AI clones - HeyGen (@HeyGen_Official) - studio quality avatars - Argil (@argildotai) - social media avatars - Overlap (@JoinOverlap) - video editor - Persona - agent builder for creators - OpusClip (@OpusClip) - video editor - Captions (@getcaptionsapp) - avatar generator + video editor To build a product: - Cursor (@cursor_ai) - AI code editor - Replit (@Replit)- automated web developer - Anychat - multi-model developer platform - Codeium (@codeiumdev) - Windsurf agentic IDE To get creative: - ElevenLabs (@elevenlabs) - text to speech - Playground (@playgroundAI) - graphics generator - Suno (@sunomusic), Udio (@udiomusic) - music generators - Midjourney (@midjourney), Ideogram (@ideogram_ai) - image generators - Runway (@runwayml), Kling (@Kling_ai), Viggle (@ViggleAI) - video generators - Krea (@krea_ai) - creative canvas - Photoroom (@photoroom_app) - image editor To learn or grow: - Rosebud (@joinrosebud) - AI journal - Good Inside (@GoodInside) - parenting app - Ada Health (@adahealth) - health management - Ash - AI counselor - NotebookLM (@notebooklm) - learn anything w/ AI - Particle (@particle_news) - personalized news To have fun: - Remix AI (@getremixai)- AI social platform - Meta Imagine (@AIatMeta) - AI-first social tools - Grok (@grok) - chatbot from xAI - Curio (@CurioBeta) - AI toys

Some key takeaways: 1. Get comfortable with looking dumb, especially in public. The best way to learn is by asking dumb questions, trying new things, and failing in front of others. You’ll uncover truths that others won’t, and more often than not, you’re not alone in having those questions. Everyone else is just too scared to ask. 2. Pair programming is a secret weapon. If you’re not doing this yet, try implementing it in your team’s workflow. It doesn’t just improve code quality; it forces intense collaboration and leads to faster unblocking. 3. Weekly updates, even if they seem like a time investment, create momentum and urgency. Parkinson’s law is real—if you set a time limit for something, it will get done in that time. Regular reviews force you to keep things moving forward, identify roadblocks early, and correct course quickly. 4. Time spent eliminating unnecessary code is never wasted. A streamlined codebase is easier to understand, less error-prone, and quicker to navigate. Make it a habit to revisit and refactor regularly, cutting away complexity. In software development, simplicity often leads to faster, more reliable results. 5. Traditional interviews are poor predictors of success. Instead: a. Use extended trials when possible b. Focus on real work product over interview performance c. Look at why candidates made career moves, not just what moves they made d. Leverage internship programs as extended interviews e. Make the interview process as close to the actual job as possible

Customer service says "Finance is facing a backlog". Of course you are - you engineered a negative opt-in and forcing people to pay for a forced upgrade and then refund them. Guys seriously - so much negative brand equity here.

Opting into @NotionHQ AI pilot has been the hugest headache every - 2nd month in a row being charged for a product which I disabled and is re-enabled on the backend. Doubling the cost of Notion with no recourse but to pay the bill and then request a refund.

How isn't important in the beginning - but what and why is more important as there are many paths to the same outcome

@lennysan @hammer_mt As a technical founder I am able to hash out 85% of the approach from high level outlines to technical load management. It's astounding.

How close is AI to replacing product managers? Closer than I expected. In collaboration with @hammer_mt (a professional prompt engineer and co-author of the O'Reilly book Prompt Engineering for Generative AI), we've begun to develop a benchmark to track how good AI is at the most common and hardest PM tasks. In our first attempt, AI beat humans at 2/3 tasks, including developing a product strategy 😮 Check out our approach, and the surpsrising results lennysnewsletter.com/p/how-close-is…

@ttunguz Verbatim what we just executed.

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. Early AI advantages have reinforced this advantage. How does this change? When AI products are sold as services, they replace in-house labor. This changes internal processes. When the internal processes change, the opportunity to replace the system of record arises because the existing workflows are no longer relevant. If paralegals’ roles and business development reps’ roles & account executives CRM reponsibilities change meaningfully, then the systems that are fully optimized to support these roles don’t match the needs of the role anymore. Systems-of-record are the horse carriages in the age of the automobile - designed for a different job to be done, outmoded, outdated. But the key is changing the internal operations of a customer, evolving the metrics, & understanding those dynamics more quickly than the rest of the market. Selling service-as-a-software enables a startup to change the way a customer works. Behavior change is hard. But successfully executed, it’s a moat. tomtunguz.com/services-as-ve…

@harleyf A lot of this boils down to old-school thinking about how you can re-use existing tracks. I would argue that the best way to execute this is to stop expanding the 401 to more and more lanes and start using some of that land to lay down the high speed track beside the highway.

I think about this daily. Who is solving this and how can I help them?

The main challenge with Consumer is moats. AI 🤖 doesn’t necessarily solve that. The key challenge is finding data lock-ins that make switching hard. B2B is a lot more natural at that because businesses have a lot of data to compute with.

@zachcoelius Just got off my 6:30 AM call with my AI team from Pakistan so yes this is absolutely 100% true.

It is clear there are two types of high performance companies emerging. 100% in person and gain the rewards of collaboration with much harder recruiting trade off Or Remote innovators who access much better talent, but a more difficult management and cohesion challenge

@jowyang A lot of that rationalization has happened at a few Telco here in Canada already. Companies like ADA I’ve been doing this for quite some time.

@adebold I think the key insight from the Salesforce announcement is that customer care and sales adr will be replaced by AI agents for level one and level two.

As AI agents replace human workers, HR will be replaced by agent orchestration software.

@jowyang just look at the investment industry and how proprietary traders and high frequency trading is largely programmatic at this point. There is a proxy or comparison to consumers shopping on the website to retail investors who will never get the equivalent value in a transaction.

Salesforce @salesforce announced AgentForce, featuring AI agents for customer care, sales, and marketing. I think that’s a first-generation use case. My vision is that each person will interface with their own primary AI agent, which will communicate with brand-owned agents. There will be no reason for humans to visit corporate websites or apps in a few years—that will be an anachronism. #DREAMFORCE2024