AI is fundamentally changing how we solve problems by forcing us back to first principles. The rapid advancement of AI capabilities isn't just changing what's technically possible - it's reshaping how users think about solutions themselves. Where we once accepted complex workflows and technical limitations as givens, AI allows us to strip away these assumptions and focus purely on desired outcomes. This shift means successful companies won't just adapt existing solutions to include AI - they'll completely reimagine what solutions can look like when unconstrained by traditional approaches.

Nobody Wants a Hammer

It reminds me of this principle from product management that says, "Nobody wants a hammer. What they want is a picture on the wall." The idea apparently traces back to Harvard marketing professor Theodore Levitt, who famously said: "People don't want to buy a quarter-inch drill. They want a quarter-inch hole." (From his work on Jobs to Be Done (JTBD) thinking.) For decades, we've built businesses around optimizing the tools - better drills, powerful bits, perfect frames. Now with AI, we can often prompt the desired outcome directly into existence. And frequently, that outcome exceeds what we could have achieved as non-expert practitioners in much less time.

Getting Out of AI's Way

Nine months ago, while experimenting with ChatGPT, I built a workout timer. I was in the practice of prompting to generate code and then copying code into my project. When I wanted to add a feature or fix a problem, I'd write a prompt and then paste the code in to be updated. As I was working on this workout timer, I went to add a feature so I wrote out the prompt of what I wanted. This time, however, I forgot to paste in my code. The code this thing responded with nailed my features but also included a user experience FAR superior to what I thought I was iterating on. I realized we weren't writing code together. I really needed to just get out of this thing's way.

Beyond Feature Addition

Even at the organizational level, every tech company needs to reinvestigate their core assumptions. I was talking to a database company about potential new features. What we came back to was that first-principles consideration of the problem. What is the future of a database? There will always be this piece of infrastructure that allows us to store, index and search information. But from a product perspective, should it remain a tool for experts wielding complex queries, or evolve into an answer machine? Traditional system boundaries are dissolving. A "database" might now handle everything from storage to analysis to insight generation - fundamentally shifting what database engineers need to optimize for. While the exact path isn't clear, one thing is: successful database companies won't just be adding AI features - they'll be fundamentally rethinking what it means to store and access information in an AI-first world.

We are in such an interesting moment. Core business objectives - delivering customer value, solving real problems - remain constant. But everything about how we achieve these goals is in flux. Even more challenging, our customers are simultaneously reinventing their own organizations, potentially transforming who our users actually are.

The Stability Paradox

The path forward requires an almost paradoxical combination of stability and radical change. By returning to first principles, we can separate the fundamental problems we're solving from the historical methods we've used to solve them. This creates the mental space needed to imagine entirely new solutions. Take customer support: the fundamental need to understand and resolve user problems remains constant, but the entire support architecture might be reimagined around AI that can diagnose issues, generate solutions, learn from every interaction and escalate only a selection of issues, allowing the team to provide a better experience across the stack. The organizations that master this balance - maintaining clarity about core problems while embracing revolutionary solutions - will win.

Fast fashion has come to software. Fast fashion brought rapid response to specific needs, ability to serve narrow market segments and the prioritization of speed over scalability to a previously staid industry. AI is bringing that to software development. I built a workout timer months back for 20 min body weight workouts while we were at an Airbnb this summer. This morning, I had Claude whip up a web app for a spinner to play Chutes & Ladders with my son. Disposable, bespoke apps for all.

There's a lot written about how everyone is now a software engineer. I've written it as well. There's nuance there. Everyone really can build now. People can speak software into existence. However, there is a lot more than that to software engineering.

Software engineering is a lot more than just getting code to run. Architecture, maintainability, upgradability, scalability, security... AI for the moment is both your friend and enemy. In August I was using ChatGPT to write tests for some code (yes, the robots are checking the robots work), and it proposed completely breaking the design of a class in service of the short term goal of getting tests to run. In its defense, it had done what I'd asked and gotten the tests to run. It takes some understanding of what you're doing to appreciate when what you've asked for and gotten from AI isn't what you actually need.

So while everyone isn't all of a sudden going to be architecting software, everyone can build. It is much easier to build software for a specific case that will be used by a relatively small community. Artisanal software, if you will. Couple weeks ago I wanted to have a visualization in the morning for my kids to see what bus we should catch to get to school in the morning. There are plenty of apps and displays for commuting, but because they are doing so much for such an expanse of people and cases, it didn't make it easy to do the one thing I needed. I wanted to see when the bus would arrive at my stop and when would that bus get to our destination for school. It took less than an hour to get something working and then another day to smooth it out for use. Cursor (the AI powered IDE) even knew how to interact with the MTA (NYC public transportation) APIs.

There will be a range of people now building small scale apps for their own use to smooth out existing rough edges. I'm hopeful that this will bring some additional polish to all the tools I'm forced to use but I'm really not the customer for. For instance, my kids' school chooses software for a range of very important reasons I'm sure, but they leave something to be desired when I consume it. I'm hopeful these companies that know they are selling software to one party (the school) that is also used by others (the parents at the school) will provide APIs and feeds for the enterprising among us to improve our own usage.

Everyone won't all of a sudden be a software engineer but many more people can now build experiences to improve their own lives. I see an explosion of targeted, niche apps and experiences on the horizon. Have you built anything fun recently?

A lot has been written about the death of product management. Let me be clear: the core of what we do will continue, while the tactical elements transform. This shift will empower those who've been focused on the craft rather than making them obsolete.

Product management connects today's activities to the future your customers want to inhabit. This means cultivating a product vision through customer conversations, competitive analysis, industry observation, and deep understanding of how customers use your product. It means connecting that vision to long and short term projects to make that vision a reality. It means assessing the impact of that work to inform future projects. And around it goes.

PMs also inevitably create mountains of documents. Clarity and alignment require documentation – product requirements, meeting agendas, incident post-mortems, email updates, presentations, and more. At its worst, this documentation burden overwhelms the more important but less urgent work of future-building.

Enter AI tools that make document creation almost trivial. While volume is now easy, the bar for quality is higher than ever. Seed AI with limited thinking, and you'll get boilerplate back. But when you've done the strategic work to understand what needs to be conveyed and why, these tools help surface overlooked elements and communicate with precision. Document format and length no longer signal competence – substance matters more than ever.

Quality and deep thinking have always been crucial, but in the past, one could get lost in (or buried under) day-to-day document generation. The core mission remains unchanged: helping teams understand what they need to do and why. That work isn't going away. Instead, AI amplifies our impact, enabling richer analysis, rapid prototyping, and compelling visuals to strengthen our case.

One final warning: don't lose the plot and use your new powers for evil. The ease of document creation isn't an invitation to bury your teams in paperwork while you elevate your focus. Being effective means figuring out what you need to say, who needs to hear it, and the most effective way to communicate it. As Blaise Pascal noted and roughly translated, "If I had more time, I would have written you a shorter letter."

It's a new world – scary, yes, but exciting. The "death" of product management might be the birth of something better.

Last week there was a call for parents to learn about a camping trip our eldest is going on with their class at school. I couldn't make the call so I asked my wife to use granola.so, an AI notepad, to transcribe and generate notes. Seeing the bulleted notes, I wondered why didn't they just email that out. I know some parents just wanted the opportunity to talk to someone about the event and to ask questions, for me though, I was happy with the bullet points. It is exciting to be in this moment where we can more easily tailor our vehicles for communication. This flexibility in formats isn't entirely new – we've seen similar principles applied in other domains.

Just as we've long manipulated image resolution on the internet, we can now adjust the level of detail in any communication – increasing specificity (upsampling) or condensing information (downsampling). Just like with images, it's much better if you start with a high fidelity asset and then translate that to formats with less and less detail, but you could nudge it up a little without too much negative impact. You can take a video communication and translate that into prose and then into tweets, if you'd like. Summarization is easy. You can take text and have the machine make it into an audio format for listening to like a podcast. You can take a podcast, extract the transcript and use gen AI to create visuals to make it into a video. The friction of transitioning between mediums is dissolving.

It's always been best practice to tailor the message to the audience and medium. Now we are only constrained by our will and thoughtfulness about how to deploy that.

This shift in communication dynamics opens up new possibilities across various sectors. In education, for instance, a single lecture could be transformed into multiple formats – detailed notes for in-depth study, concise summaries for quick review, or even interactive visual presentations for different learning styles. In business, complex reports could be efficiently condensed for executive summaries or expanded into comprehensive presentations, catering to diverse stakeholder needs.

Ultimately, how we communicate fundamentally remains the same – we still aim to inform, persuade, and connect. But the versatility in tailoring our messages has been forever changed. As we move forward, our challenge lies not in the technical ability to transform our communications, but in our thoughtfulness and skill in leveraging these tools effectively. The future of communication is not just about what we say, but how we choose to say it across an ever-expanding array of formats and mediums.

On this last day of my challenge to publish 30 pieces of writing in 30 days, I'm feeling reflective. The key to making this challenge both navigable and helpful was focusing on the process, not the result. My goal was simple: publish every day. It didn't matter what I wrote about or if it was any good. I just needed to get comfortable pushing that publish button. This approach unburdened me and, ironically, gave me greater opportunities for success than obsessing about success ever would have.

Why Process Trumps Perfection

I want to get better at writing. I want to create content that people find interesting and valuable. But if I'd set out to write something compelling every day, it would have been paralyzing. I'd have been overly critical, over-edited, and produced very little.

Writing more is a necessary step towards gaining proficiency and giving myself the opportunity to write something good. By mandating that I work the process, I put myself in a position to achieve the higher goal. Focusing solely on that higher goal wouldn't have gotten me there.

You might wonder: Why do I have to play this game with myself? Aren't I smart enough to deconstruct the steps needed to achieve the end result? Maybe. But generally, no. For me, it's almost like only by abandoning the higher goal could I get closer to it.

The Sustainability of Process-Focused Approach

Trying hard to achieve something is great. It's noble. There's no shame in it. But it's not reliable or sustainable. Dedicating oneself to a process, however, is a much better pathway to long-term goal achievement. And not just achievement once, but on an ongoing basis.

Learning from the Pros

I remember listening to a podcast that discussed how Taylor Swift and Frank Ocean are hit-making machines. The reason isn't just that they try hard. They've dedicated themselves to honing their craft, constantly refining their process of creation. This builds a machine of success.

Similarly, I recently listened to Lenny Rachitsky interview Nikita Bier about his "playbook for winning at consumer apps." Bier emphasized that success in consumer social apps is capricious. What he and his team excelled at was the process. They developed a systematic approach: target a single community with an app, measure its performance, track usage sharing within the community, then measure if it spread to other communities. This refinement of process puts one in the zone to be successful.

The Takeaway

You can't just want six-pack abs. You have to love each and every day in the gym. Ironically, focusing on progress rather than perfection is what brings you closer to perfection. Keep taking swings. The harder and more consistently you work, the luckier you get.

Ultimately, it's the dedication to process that opens the door to achieving our goals consistently over time. This approach often leads to success in ways we couldn't have predicted or forced. So when you're staring down a big goal, try shifting your focus. Instead of fixating on the end result, pour your energy into refining and showing up for the daily process. The results might surprise you.

I have a computer science undergrad. I learned to write C in college. I worked as a software engineer for a few years after school and I've worked in the software space my whole adult life. Over the years, I've built personally or as a part of a team lots of things. Most recently I've been experimenting a lot with LLMs first to build software again and now to do some writing. Coding with the assistance of LLMs has been incredible. Writing with LLM assistance hasn't felt like the same order of magnitude improvement but it has also been transformative. My next project involves me building on top of LLMs and I expected the leap in capability around natural language interaction but I wasn't prepared for the implications to my software stack and how it's tuned.

My first project I'm undertaking involves using LLMs as a part of the infrastructure to synthesize natural language communications and perform transformations on it. I have built software before and I understand that process. One of the comforting things about code is that given the same inputs, it will always generate the same outputs. If there's a bug in the code, in an odd way, you can count on that bug. If you understand the circumstance that trigger the bug, you can reproduce it consistently.

Working with an LLM is putting this probabilistic black box into the mix that truly enables natural language processing that is unparalleled. Doing that piece in another way would be, for a small team, essentially impossible. But now we have these APIs and it's really blowing my mind a little as I try to operationalize it for the first time. The prompting is code and configuration. The benefit of prompting is that we can use natural language to interact and direct the LLM. It's easier. The challenge is that natural language is so imprecise.

I'm early yet so I don't have takeaways or recommendations but it's hard not to wander about how this progresses. Early ChatGPT days I thought "prompt engineering" was too cute by half. Now, as I continue to fine tune the proper incantation that can refine my desired outputs, I have to think maybe there's something to it.

I'm just getting started but as I dive deeper into LLM-integrated development, I'm realizing this isn't just a new tool—it's a fundamentally different approach to building software. The shift from predictable code to probabilistic outputs is forcing me to rethink how I design, test, and debug systems. Honestly, I have far more empathy now for all of the widely publicized missteps of the big players in operationalizing these technologies. While the potential is exciting, the challenges are significant. We'll need to develop new methods for ensuring reliability and consistency in AI-augmented systems. It's still early days both for me and this tech. But one thing's clear: the skills that make a good builder are evolving, and we'll all need to adapt.

As I've written before, I've found using an LLM as a writing buddy to be incredibly helpful. The feedback on the flow of my writing, transitions, intros and conclusions has really helped me produce and publish on a schedule that I just wouldn't have been able to before. To be fair, sure I could have published if I had no standards. And maybe in the future after I've learned from its coaching I could. But starting from zero to daily, it's enabled me to publish at a pace and quality I wouldn't have been able to achieve otherwise. However, it seems as if I'm most productive when I understand enough about what I want to create that I can explicitly guide it toward success. When I don't know how to do that, it can feel like I'm just shaking a magic eight ball repeatedly and hoping for my preferred answer to show up.

I'm well into my 30 days of publishing and I went into the archives for what I might publish in the future. I found a little narrative I'd written about a cascade of parental drama 5 years ago trying to get my 2 young kids into the bath tub. I asked Claude for advice using my existing prompt for writing feedback and I did not like the feedback I got back. Not that it hurt my feelings but the recommendations didn't seem to improve my story. Reflecting on this experience, I considered several explanations for the LLM's limitations.

One possibility is certainly that I haven't figured out how to use the LLM effectively to set it up for success in helping me improve this type of writing. That's almost certainly true to some extent. It was my first attempt. However, another possibility is that it's much harder to explain the type of help you want in something that's more subjective. I circled back to try again and the results were a little better but it still didn't feel as powerful as in other scenarios.

My theory is that based on how LLMs work, essentially probabilistically guessing at what a good answer might be, it defaults to offering up the most popular styles that your submission can fit into. If the NY Times represents the majority of it's training data, it's going to try and help you by working your submission into the form of the NY Times. You can get specific and make adjustments but you need to language and understanding to guide it.

If you are knowledgeable enough about what you are trying to create, you can speak to the sub-components that contribute to the desired effect. If you lack that language, it's very hard to guide it.

Capable people are able to execute at scale on an unprecedented level now. Less capable people will be able to very quickly ape familiar styles but it's harder to create something uniquely powerful without an understanding of and the language to articulate the elements that contribute to the desired creation.

As we continue to work with LLMs, it's crucial to recognize their role as skill magnifiers. They can supercharge productivity for the proficient and provide a solid foundation for novices. But their true potential lies in accelerating our learning and skill development.

The challenge lies in guiding these AI tools effectively, especially in creative realms. It requires not just technical know-how, but the ability to articulate nuanced, subjective qualities.

As creators and builders, our task is clear: leverage LLMs where they excel, but more importantly, develop the expertise and language to guide them effectively, especially in creative and subjective tasks. The future belongs to those who can seamlessly blend AI assistance with human creativity, insight, and the ability to provide nuanced guidance.

What's your take on this balance? How do you see it shaping the future of content creation?

Web3 platforms like Farcaster excel in areas that are revolutionary from a technological and ideological standpoint. However, these strengths often don't align with the immediate priorities of mainstream users:

• Web3 Platforms Offer:

  • Credible neutrality

  • Composability

  • Potential for creator rewards

• Mainstream Users Prioritize:

  • Entertainment

  • Information

  • Engagement on their posts

Web3 has a lot to offer in as much as it can create a better experience, but mainstream audiences want those better experiences. They aren't interested in learning about the principles and investing in what might be. They just want to be where everyone else is. While these differing priorities create a fundamental challenge for Web3 platform adoption, they also intersect with a more nuanced issue: the evolution of social currency in the digital age.

Traditional social platforms have long relied on likes and interactions as a form of social currency. Web3 introduces a new dimension by bringing real and imaginary financial elements into the mix. However, striking the right balance is crucial. Monetary gain must remain secondary to primary social interactions, or the platform risks losing its core value.

If users feel that others are interacting solely for financial gain, it can degrade the quality and authenticity of social interactions. This is a balance that many platforms in Web3 struggle to maintain.

Many Web3 social platforms currently lean more towards "Fi" than "Social." They essentially function as finance games wrapped in social trappings, where the financial aspect becomes the primary focus. While this can be entertaining for some, it's not a sustainable model for broad adoption.

For these platforms to achieve lasting success, they need to attract and retain users who are primarily there for the social aspects, with the financial elements serving as an enhancement rather than the main draw. This shift in user base and motivation is crucial for long-term viability.

To bridge Web3 experiences to mainstream audiences, platforms need to rethink their approach:

1. Lead with traditional social currency, not digital tokens. The primary reward for participation should be social engagement and recognition.

2. Create experiences that are inherently engaging without relying on financial elements. The core functionality should be satisfying even if all token-related features were removed.

3. Implement what could be called a "Social in the streets, crypto in the sheets" approach. This means designing for satisfying social experiences first, then integrating crypto elements subtly to enhance, not dominate, the experience.

The key is to develop an experience that doesn't exist primarily for the blockchain but is made better because of it. Users should be able to enjoy the platform fully before they even discover that wallets or tokens are involved. While these bridging strategies offer a general framework for Web3 adoption, applying them to specific platforms reveals unique challenges and opportunities. Let's examine how these principles might play out in Farcaster's case, and what that could mean for its future trajectory.

I don't presume to have this figured out, but as framed, it's possible that the wide ranging appeal we see in Web2 is just not a possibility for a Farcaster client. The benefits of it may require some of the friction. Currently the cost of Farcaster usage is about $1 USD/month. This includes gas fees for account creation as well as for storage social data like posts, reactions and profile info. The fee also serves as an impediment to spam. It just can't be credibly neutral without the decentralization and the resulting fees.

The advertising supported business model allow for lower friction of use potentially allowing those platforms to be everything to everyone. Individuals using the application aren't paying for the service but instead the advertiser is in order to reach those individuals. Mainstream audiences don't care about how a service is provided. They don't want to have to learn. So while crypto natives commonly understand the nature of open protocols and digital ownership and the costs they include, someone like my mother or sister is not and they aren't trying to learn about it. A Farcaster client could sponsor it themselves but only if they have some other business model that makes that possible.

Success for Farcaster and its clients just may look different than the social networks that have come before it.

While this discussion focuses on Farcaster, the general challenges it faces are relevant to any project trying to appeal to a broader audience. The key to mainstream adoption of Web3 technologies lies in emphasizing the value desired by people who aren't inherently interested in the technology for its own sake. At the same time, blockchain products cannot compete directly with their Web2 predecessors. They are different animals with different strengths and weaknesses. As builders and consumers, we have to remember that as we define what success looks like.

As a space, we've certainly made a lot of progress recently. Coinbase's smart wallets and Privvy's embedded wallets have helped many projects diminish friction in their experiences. There are many examples of big improvements.

• Coinbase's free USDC transfers on the Base layer 2

• MoshiCam's integration of the Coinbase smart wallet and the Farcaster network

• Zora's creator focused marketplace and new mobile app

• Warpcast and all the Farcaster apps

We see these as the great achievements they are because we know what came before. Mainstream audiences don't care about what came before. They only care about what is today and we still aren't there yet.

Future success in this space depends on creating experiences that are inherently engaging and satisfying, with Web3 elements serving as enhancements rather than the primary focus. By leading with user-desired value and integrating blockchain features seamlessly, platforms can bridge the gap between Web3's potential and mainstream adoption.

• AI is a great writing copilot on making sure you are conveying what you want via your intros and conclusions. Let it.

• Back to School with AI

  • Upload pictures or PDFs to manipulate or export structured data

  • Get school contacts into your phone easily

  • Summarize the torrrent of school emails

• Side Hustle Manifesto: A personal journey through coding, writing, and building.

• Can Google and Gemini Catch Up to Their Own Potential: My frustrating encounter with Google's Gemini AI in Google sheets reveals a gap in their operationalizing AI for their customers. UPDATE: Since I wrote this, I see Google has added gen AI editing of images in Google Photos. I haven't played with it yet but I'm hopeful.

• Artifacts Everywhere!: Claude's deployment of Artifacts to all users is an example of AI done right. The ability to preview charts and code in the Claude experience helps the magic come alive immediately for users.

• The Second Best Time To Start A Writing Practice Is Today: Overcome perfectionism, leverage AI, and start writing today. In an era of accessible content creation, your audience becomes your key differentiator.

Have a great week!

Some view AI assistance as "cheating," assuming it makes producing quality content effortless. This perspective misses a crucial point: excellence in any field—be it gymnastics, ice curling, or content creation—always demands hard work. AI simply shifts the nature of that work.

When we encounter soulless, overlong writing, it's tempting to blame the AI, but that's like faulting a pen for poor penmanship. The real issue lies with the creator who deemed such output acceptable. It's reminiscent of those who discovered PowerPoint transitions and proceeded to obscure their message with flashy effects. It's easy to get distracted by the allure of new technology. The core challenge remains: conveying a message effectively.

The backlash against AI-assisted content likely stems from a sampling bias. We notice the poor examples because they stand out, while skilled AI users fly under the radar. This skews our perception of AI's impact on content quality.

There's no definitive answer to what's appropriate here. Only society will decide over time. However, I'd bet on widespread deployment and acceptance in the future. The only question is how quickly we'll get there.

As we navigate this transition, let's focus on the fundamentals: clear communication, purposeful creation, and measuring our output by its effectiveness in conveying our intended message. AI is just another tool—it's how we use it that truly matters.

The benefits of writing are well-documented and compelling:

• Improved clarity of thought

• Stress reduction

• Enhanced communication skills

• Greater ability to tell a compelling and persuasive story

These aren't just anecdotal claims. A study published in the Journal of Experimental Psychology found that expressive writing can improve working memory, which is crucial for cognitive tasks and managing daily stressors [1].

Over three weeks ago, I set myself a challenge: write and publish 30 pieces in 30 days. The aggressive timeline was intentional, forcing me to leverage AI tools and overcome perfectionism. This is day 24, and the experience has been transformative.

This daily publishing requirement has been pivotal in conquering perfection paralysis. Previously, I'd agonize over every word, seeking feedback and making endless revisions. The result? Months between posts. And for almost no one to end up reading it. Case in point: my first post here was on April 15th, and it took until August 15th to post the next one as part of this challenge. Progress, not perfection. Consistency, not intensity. This is a marathon, not a sprint.

The practice has also sharpened my ability to generate ideas. I now maintain a notes folder full of potential topics, jotting down thoughts as they come. It's similar to how carrying a camera everywhere trains you to see the world differently, always on the lookout for captivating scenes. This writing mindset has benefits that extend beyond the act of writing itself.

So, what's holding you back? If you're worried about the quality or reception of your writing, let me ease your mind with some hard truths. Based on my own experience with this challenge, engagement with online content when you start is often quite low. Even for posts I considered "popular," I found that most readers only skimmed a small portion of the content.

The takeaway? In the beginning, hardly anyone cares what you write. This is actually liberating. You have the freedom to fail in relative privacy, find your voice, and refine your message without the pressure of a large audience.

Think of your early writing as brand advertising. Marketing research shows it takes multiple exposures before someone takes a message or messenger seriously [2]. I've experienced this myself with Instagram purchases, often ruminating on a product after several exposures before committing. You may need to have a real practice and body of work before people start catching on to what you're putting down.

If you're inclined to start, my recommendation is simple: begin publishing today. Set a challenge if it motivates you. I told a friend I'd pay him $100 if I didn't meet my 30-in-30 goal. In an AI-driven future, building an audience will likely become even more valuable. So share whatever speaks to you - writing, poetry, visual art - anything that gets you going.

Remember, the future of digital content is being shaped now. Your unique voice and perspective are valuable. The audience you build today could be your greatest asset tomorrow. So don't wait for the perfect moment - start writing, start sharing, and start building that audience now.

[1] Klein, K., & Boals, A. (2001). Expressive writing can increase working memory capacity. Journal of Experimental Psychology: General, 130(3), 520-533.

[2] Krugman, H. E. (1972). Why three exposures may be enough. Journal of Advertising Research, 12(6), 11-14.

But what exactly are artifacts? Artifacts are dedicated windows that display substantial, standalone content generated by Claude in response to user requests. They allow users to view, edit, and interact with AI-generated outputs directly within the chat interface, rather than having to copy and paste content into separate applications. Importantly, when possible, artifacts display a preview of the result of any code generated, rather than just the code itself. This means you can see the output or functionality of the code without needing to run it in a separate environment.

This feature represents a significant leap forward in how we interact with AI. The true power of artifacts becomes apparent when you start using them. For instance, asking Claude to build a simple tic-tac-toe game results in a playable game right there in your browser. It's a stark demonstration of how AI can move beyond text generation to create interactive, functional content on demand.

When it comes to working with LLMs, I'm not particularly concerned about which specific model I'm using. For those building AI-powered applications, there might be reasons to prefer one over another - I've heard arguments for ChatGPT's API and programmer interface superiority, and there are certainly considerations around optimizing price versus performance in using older models. But for my purposes, I simply use the latest available model.

What's really shifted my usage patterns is Claude's Artifacts. It's the primary reason I've moved away from ChatGPT and haven't given Gemini much consideration. The ability to generate and interact with content directly within the chat interface is transformative.

Experiencing Artifacts firsthand makes you want them integrated everywhere. Consider the possibilities:

• Enhanced blogging: Imagine incorporating working demos directly into blog posts, allowing readers to interact with the concepts you're discussing.

• Dynamic emails: Proposals could include interactive charts that recipients can manipulate, making your communications more engaging and informative.

• Interactive presentations: Talks and slides could come alive with real-time data visualizations, significantly improving audience engagement.

LLMs are already revolutionizing data manipulation, prototyping, and demo creation. Artifacts take this a step further by making the output immediately interactive and useful.

As more people leverage these tools, it's likely to raise the bar on what we expect from content creation and consumption. The ability to include interactive charts, graphs, and demos in various forms of communication could become the new standard.

While Artifacts are currently a standout feature of Claude, the concept they represent - AI-generated, interactive content - is likely to become more prevalent across various platforms and tools. As this technology evolves, we can expect to see even more innovative ways to create and share interactive content.

In conclusion, the introduction of Artifacts represents a significant step forward in making AI-generated content more tangible and immediately useful. It's an exciting development that hints at a future where the line between content creation and interactive experiences becomes increasingly blurred.

1. Turn Paper into Pixels
It works for digital paper as well. Upload a PDF or a picture of a document and ask AI to create a sortable table for easier viewing or to copy elsewhere.

2. Get School Contacts Into Your Phone
Remember that email with 50 staff contacts? Copy, paste into an AI, ask for a vCard file. Save the contents (`contacts.vcf`). Import into your phone. I'd much prefer to search my phone contacts for the email address to alert that my child will be home sick rather than digging through emails.

3. Summarize School Emails
Every school email is like a recipe on the internet. You gotta hear about the hike in the Andes when they first tried this dish when all you want to know is how many cups of flour you need. Your mileage may vary here so I recommend double checking the important bits but this is still helpful for triage in the midst of running around.

Here's to another year of organized chaos. May your coffee be strong, your patience endless, and your AI assistant eerily efficient. Best of luck.

1. Write what you want to say

2. Ask AI what you're trying to say

3. Update your intro to tell people what AI said you're going to say

4. Update your conclusion to tell people what AI said you said

It's like AI inception of your own thoughts! 🤖🧠✍

Many people predict that AI will be a space where the rich get richer. Incumbents have the data and the capital to do the expensive work of training models. They also have the applications that AI can be applied to. Startups always have a shot. They move fastest. The AI powered IDE, Cursor, is pretty much all anyone can talk about right now. But what does this mean for everyday users like us?

I wouldn't hate it if I get some of that AI magic in my Google Photos, right? If my existing Google Docs files were able to leverage this technology without me having to copy and paste from Claude, that'd be pretty nice. And who better to deploy this than Google who pioneered much of the research that has led us here. This is supposed to be Google's real edge against the competition, but I'm not sure it's going so well.

When AI Meets Real Life: A Tale of Spreadsheet Woes

Let me share a recent experience that highlights the current state of affairs. With the start of the school season, my son's school sent parents a PDF document that listed what advisor group each kid would be in for the 5th grade. I scrolled down and found my son and a couple of his friends manually but I wanted to see everyone on the list in their groups. I really just wanted to sort the list. I couldn't copy-paste the doc into Sheets for some reason. Some times it works. Some times it tries to plop the whole file into a single cell. Today was the latter. I had Claude create a table out of it that I was then able to copy and paste into Sheets. Sorted it easily and saw the group.

Then I wanted to color each row by the group so I could visibly see the pattern even when not sorting by the group. Gemini was right there on the side. So I asked it to help. It was NOT helpful. Not even a little.

It seems as if Google is currently using the Gemini branding to respond to queries and perhaps to generate images but it can't yet actually execute tasks in their application. To be fair to Google, they have many many complicated and critical applications. Deploying gemini as a bolt-on beside the application was certainly the easiest first step. However, it doesn't realize any of the promise of Google applying AI tech to their data and products for amazing results. It's exceedingly underwhelming. Right now it seems to be wizard for creating templates of tables in Sheets.

When Claude Saves the Day (Sorry, Google)

Frustrated with Gemini's lack of help, I decided to take matters into my own hands. I tried to do it manually on my sheet and went to the "Conditional Formatting" option to tell it to color each row based on the advisor but when I realized I had to execute multiple clicks to even enter the data and hardcode the advisor names. I gave up. Frustrated with Google as a whole, I popped back over to my Claude window that converted the PDF to a table I could copy and asked it: Can you make it sortable and color each row by the value in the advisor column? And there I had what I needed. Claude allowed me to build a custom web page to visualize this info quicker than using Google's LLM to interact with a Google product. This stark contrast really drives home the point about Google's current AI operationalization shortcomings.

The Race Is On: Google's AI Growing Pains

This is clearly a moment in time. I understand from others that Gemini is really quite powerful. I know the teams there are very smart and I imagine they are just working under constraints I don't know about. That's likely what created such a disappointing result. However, if Google doesn't address this gap soon, other models and experiences are going to capture the layer of usage that involves the most simple and basic data manipulation we all engage in daily. Google Sheets did this to Excel. Now every LLM is going to do this to Google.

For now, I'm sticking with what works best. Doing it in Claude was just too easy. The real question is: Can Google with all of their challenges of incumbency realize their potential to create a game-changing experience for users combining their products and the tech they pioneered. I hope they figure it out though. For their sakes, and for ours as everyday users who could benefit from truly smart, AI-enhanced tools.

The Dawn of Unprecedented Opportunity

The internet created this equally accessible frontier where anyone could throw up a web page and serve the world. It could be a media company, a store front, a game, whatever. The smart phone put the internet in our pockets adding the gps and cameras to our internet interactions. Cloud computing brought in an era where you didn't have to manage racks of servers in order to scale an internet business. SaaS products have given us a range of applications that used to be whole departments at larger organizations that any little startup can now leverage and compete on equal footing. AI has now supercharged the individual enabling us to write code, copy write, design, and a host of other things at a boosted rate of productivity.

As we've seen, the technological landscape has evolved dramatically, creating a fertile ground for innovation. But how does this translate to individual opportunities?

The AI Revolution: Empowering the Individual

As I said, there has never been a better time to build something. The breadth of things an individual or small team can execute on is historically unparalleled. The smaller the team, the smaller the opportunity they can consider that a larger operation wouldn't find attractive. Even if you don't know what you might want to build, the ability to LEARN is similarly unparalleled right now. An LLM can provide very adequate tutoring and feedback on a range of things enabling us to level up in all dimensions.

With these powerful tools at our fingertips, the question becomes: how can we harness them to bring our ideas to life?

A Personal Journey: Rediscovering Coding

To start my journey, a few months ago I started trying to see how AI crutches could help me to write code again. I have a computer science undergrad and I worked as a software engineer decades ago but I learned to write C in school and professionally I wrote Java and Ruby for a short bit. I have a lot to catch up on. I built smart contracts for digital collectibles and a Next.js typescript app to enable people to interact with those collectibles and contracts. It's not live yet but I'm going to pull the trigger really soon.

It was an amazing experience. It was the first part of my journey so progress didn't follow a straight line. I couldn't plan what I'd be building because I didn't know what I'd be able to build. There were many cases where after seeing what I'd built, I realized that it wasn't the best experience for what I wanted to build. At times I had incredibly satisfying bursts of productivity where it felt that things would just spring forth from my mind into reality. Other times, I might be stuck for a few days as I needed to work through the assorted pieces of my app I was co-creating with AI in order to debug a problem. At one point I realized my app had night mode, meaning it had an alternate darker color scheme for night time vs day time, and it wasn't something I'd explicitly tried to create. As it turns out, yes I could write code again. It's been incredibly empowering.

While my coding journey was exhilarating, I realized that building isn't limited to software development. There are countless ways to create and innovate in this new era.

Experimenting with Quick Builds: The Workout Timer

As a result of all of that learning I'd put together the foundations so that one day I built a workout timer in the space of 25 minutes so that I could have the appropriate intervals displayed on my computer as I moved through the workout. I'd already figured out the basics of what language I could write in, JavaScript, framework to build with, Next.js, CSS tooling to use, Tailwind and where to host it, Vercel. Over the next couple days I polished the timer app and then I manually hosted a fitness fight club with some friends. Everyone put in $20. We all had to do the daily prescribed workout on my app for 5 of the next 7 days. If you did your workout, you got your $20 back. If you didn't, your buy-in was split amongst those that did. Despite the fun (and the $5 bucks I earned off a friend) it didn't make sense to scale up into a larger experience.

This experiment, while fun, taught me valuable lessons about rapid prototyping and market testing. But I didn't want to limit myself to just one form of creation.

Embracing the Writing Challenge

I realized I was leaning deeply into building but I also wanted to flex other creative dimensions. I committed to publishing a piece of writing every day for the next 30 days. This is my piece for day 20. I wrote a little bit about how I've used Anthropic's Claude as a writing partner and it's been eye opening. The necessity of publishing every day has been a helpful forcing function in getting me to get my ideas out there and to not be too precious about it. It's been a little challenging in the last couple days as my family and I are moving back into our house in Brooklyn after a year abroad but even with that, this is the point. I'm leveraging AI tooling to produce in a way I absolutely could not without them.

After this 30 day challenge is over, I hope to continue this new writing habit but at a less intense rate. I'd like to publish maybe 1-3 times a week but I know that doing something every day is in some ways easier for me than trying to do it a couple days a week with a loose schedule. I'll figure it out. I don't want to lose this muscle.

As I push through this writing challenge, I'm constantly reminded of the power of consistency and the tools at our disposal. But what comes after these 30 days?

Looking Ahead: The $10 Internet Challenge

The next stage of my journey will be to make $10 bucks on the internet off of software. Consulting would be cheating. I need to find some niche problem that I can solve and charge for. I've clearly built lots of software over the years and solved many problems, but I've never been this agnostic as to the type of problem. I'm not bound by an existing company or area. I just need a problem software can solve that someone is willing to pay $10 for. Or ten people are willing to pay $1 for. Or I can give 50% off the $20 price tag. Regardless, I'm trying to put together my toolkit to see what I can offer the internet.

This next stage of my journey represents the essence of what I've been exploring: the ability to identify problems, create solutions, and deliver value in the digital age.

Conclusion: Seizing the Moment

There has never been a better time to build. I truly believe that we are in a moment where an individual or small team can punch above their weight drastically. If you have the will to learn and build, that can be you and me too.

In this era of unprecedented opportunity, the only limit is our imagination and willingness to learn. Whether it's through coding, writing, or solving unique problems, there truly has never been a better time to build. What will you create?

In the world of retail, keeping an accurate picture of inventory where theft, damage, loss and other unexpected elements can be difficult. In agriculture, weather, disease, and pests can skew your numbers. In construction, unexpected site conditions, material delays, and the inherent variability of human labor can make it challenging to maintain an accurate digital representation of project progress and resource allocation. The ongoing record keeping to maintain an accurate digital view of the world creates additional overhead to leverage technology.

Given these challenges, is there a way to bridge the gap between the physical and digital worlds without adding significant overhead? If it no longer needs to be ancillary work that an otherwise useful person needs to do but is instead a task that can happen automatically, does that enable this space to become more digitally trackable? In the construction example for instance, could off the shelf web cameras stream worksites video that AI then processes? Could this enable more easily generated progress reports and estimate progress or perhaps even create an alert to potential structural issues that may be evident in photographic evidence but not as apparent to someone on the ground due to the gradual nature of the changes? Certainly, AI query prices would need to come down but they are already trending in that direction and perhaps greater value could be seen if the data gathering were more persistently possible.

As industries like construction, retail, and agriculture continue to grapple with the physical-digital divide, AI may offer a path forward. The question remains: can AI truly bridge this gap and provide tangible value, or will it become another overhyped solution in search of a problem?

Sounds unconventional, right? That's exactly the point. Adoption of AI tools in the workplace generally sits at the individual level rather than the organizational level, and it's unevenly distributed. This role-swapping hackathon idea isn't just about mixing things up—it's a strategic approach to address the challenges of widespread AI adoption.

By pushing people out of their comfort zones and into unfamiliar territory, we create an environment where experimenting with AI tools becomes not just beneficial, but necessary. This approach provides both a concrete task and a compelling reason to explore new methods, potentially accelerating AI adoption across the entire organization.

To contextualize this, I’ve spoken to many people who say they want to play with AI but they don't seem to manage it. Maybe they tried one prompt and then walked away. They aren't sure how to operationalize it.

The thing I have seen be successful is to have a project. If you have something you want to build, you now have an objective so it isn't just learning for its own sake. Given space to work on a project but without their usual tools, they have greater reason to experiment with new tools. Anyone trying to pick up a new way of doing something, even if that way is objectively better, is likely to have an initial drop in productivity before they find their stride. In this scenario with traded responsibilities, people can easily understand that doing it the old way isn't even a possibility so they HAVE to find a new way.

I've heard that even in the areas where people are bringing new AI tools into work, there's very little knowledge sharing for fear of how it will be perceived or if it will be explicitly allowed by the organization. Clearly endorsing experimentation to achieve greater efficiency and publicly applauding those gains enables the organization to recognize that this is endorsed and the organization as a whole can reap the benefits.

In conclusion, the key to operationalizing AI tools in the workplace lies in providing employees with two critical elements: a concrete project to work on and the space and encouragement to approach it in novel ways. The role-swapping hackathon I've proposed is designed to deliver both of these elements in a unique and engaging format.

By pushing teams out of their comfort zones, this approach:

1. Encourages hands-on experimentation with AI tools

2. Fosters cross-functional understanding

3. Breaks down barriers to AI adoption

4. Promotes organization-wide knowledge sharing

Moreover, by explicitly endorsing and celebrating AI experimentation, organizations can create a culture that embraces innovation and reaps the full productivity benefits of these powerful new tools.

While this hackathon idea offers one path forward, I'm curious to hear about other strategies. Have you encountered or implemented other effective ways to encourage an entire organization to harness the productivity benefits of AI?

I haven't been through the list exhaustively but the thing I'd add is Perplexity. I dropped his post into Perplexity and asked it to give me the urls and descriptions of all the apps mentioned. Voila!

Programming Tools

1. Cursor AI: AI-powered IDE built on VS Code with advanced code assistance. https://www.cursor.so/

2. Replit: Browser-based collaborative coding platform. https://replit.com/

3. Claude 3.5: Advanced AI language model by Anthropic for various tasks including coding. https://www.anthropic.com/

Design Tools

1. V0: AI-powered design tool for creating and editing web sites. Generate UI with shadcn/ui from simple text prompts and images. https://v0.dev/

2. Playground AI: AI image generation and editing platform. https://playgroundai.com/

3. Galileo AI: AI-assisted UI/UX design tool. https://www.usegalileo.ai/

Copywriting Tools

1. Jenni AI: AI-powered writing assistant for various content types. https://jenni.ai/

2. Copy AI: AI copywriting tool for marketing and content creation. https://www.copy.ai/

3. Anyword: AI-powered copywriting platform with performance prediction. https://anyword.com/

Product Management Tools

1. Productboard: Product management platform for prioritizing features and roadmapping. https://www.productboard.com/

2. Aha!: Product roadmap and strategy software. https://www.aha.io/

3. Airfocus: Modular product management platform. https://airfocus.com/

Lead Generation Tools

1. Gamma AI: AI-powered presentation and document creation tool. https://gamma.app/

2. MailerLite: Email marketing and automation platform. https://www.mailerlite.com/

3. Systeme.io: All-in-one marketing platform for online businesses. https://systeme.io/

Video Production Tools

1. Revid: AI-powered video editing and creation platform. https://www.revid.ai/

2. Pictory: AI video creation tool from text and images. https://pictory.ai/

3. Vidyo AI: AI-powered video editing and repurposing tool. https://vidyo.ai/

Growth Hacking Tools

1. MutinyHQ: AI-powered website personalization platform. https://www.mutinyhq.com/

2. GrowthBook: Open-source feature flagging and A/B testing platform. https://www.growthbook.io/

3. Voyantis: AI-driven user acquisition optimization platform. https://www.voyantis.ai/

4. Phlanx: Influencer marketing and social media analytics platform. https://phlanx.com/

5. Viral Loops: Referral marketing and viral campaign creation platform. https://viral-loops.com/

Data Science Tools

1. Obviously AI: No-code machine learning and predictive analytics platform. https://www.obviously.ai/

2. H2O.ai: Open-source machine learning platform. https://h2o.ai/

3. DataRobot: Enterprise AI platform for building and deploying machine learning models. https://www.datarobot.com/

4. BigML: Machine learning platform for predictive analytics. https://bigml.com/

5. RapidMiner: Data science platform for analytics and machine learning. https://rapidminer.com/

Sales Tools

1. Apollo.io: Sales intelligence and engagement platform. https://www.apollo.io/

2. Outreach: Sales engagement platform for managing customer interactions. https://www.outreach.io/

3. Gong: Revenue intelligence platform for sales teams. https://www.gong.io/

Customer Support Tools

1. Intercom: Customer messaging platform for support, marketing, and sales. https://www.intercom.com/

2. Front: Customer communication hub for teams. https://front.com/

3. Chatdesk: AI-powered customer support platform. https://chatdesk.com/

Financial Analysis Tools

1. Finta: Financial modeling and analysis platform for startups. https://www.finta.io/

2. Runway: Financial planning and analysis software for startups. https://www.runway.com/

3. Pulse: Cash flow management and forecasting tool for small businesses. https://pulseapp.com/

Legal Assistant Tools

1. DoNotPay: AI-powered legal services app for consumers. https://donotpay.com/

2. LawDepot: Online legal document creation platform. https://www.lawdepot.com/

3. WonderLegal: AI-powered legal document creation platform. https://wonderlegal.com/

4. Spellbook: AI-powered contract drafting and analysis tool. https://www.spellbook.legal/

Recruitment Tools

1. Dover: AI-powered candidate sourcing and outreach platform. https://www.dover.io/

2. Fetcher: AI-driven recruiting and candidate sourcing platform. https://fetcher.ai/

3. Gem: Talent engagement and recruiting analytics platform. https://www.gem.com/

Marketing Strategy Tools

1. SEMrush: All-in-one digital marketing toolkit for SEO, content, social media, and more. https://www.semrush.com/

2. Ahrefs: SEO tools and resources for marketing research and analysis. https://ahrefs.com/

3. Moz: SEO software and tools for improving search rankings. https://moz.com/

UX Research Tools

1. Maze: Rapid testing platform for user research and product teams. https://maze.co/

2. Hotjar: Behavior analytics and user feedback platform. https://www.hotjar.com/

3. UserTesting: Platform for gathering user feedback and conducting usability tests. https://www.usertesting.com/

Business Analysis Tools

1. Holistics: Self-service analytics and SQL-based reporting platform. https://www.holistics.io/

2. Metabase: Open-source business intelligence and analytics tool. https://www.metabase.com/

3. Preset: Cloud-native analytics platform built on Apache Superset. https://preset.io/

Project Management Tools

1. ClickUp: All-in-one productivity and project management platform. https://clickup.com/

2. Notion: All-in-one workspace for notes, tasks, wikis, and databases. https://www.notion.so/

3. Coda: Doc-based platform combining documents, spreadsheets, and applications. https://coda.io/

4. Linear: Software project management tool for high-performance teams. https://linear.app/

Public Relations Tools

1. Prezly: PR software for storytelling and media relations. https://www.prezly.com/

2. OnePitch: PR platform connecting brands with relevant journalists. https://onepitch.co/

3. Prowly: Media relations management and storytelling platform. https://prowly.com/

Social Media Management Tool

1. Zaap AI: AI-powered social media management and content creation platform. https://www.zaap.ai/

Video Production Tools

1. Runway: AI-powered creative tools for video editing and generation. https://runwayml.com/

2. Descript: All-in-one audio/video editing and transcription platform. https://www.descript.com/

3. VEED: Online video editing and creation platform. https://www.veed.io/

Email Marketing Tools

1. beehiiv: Newsletter platform for creators and publishers. https://www.beehiiv.com/

2. Substack: Platform for subscription newsletters and podcasts. https://substack.com/

3. ConvertKit: Email marketing platform for creators. https://convertkit.com/

Idea Generation Tools

1. GummySearch: Reddit-focused market research tool. https://gummysearch.com/

2. Treendly: Trend discovery and analysis platform. https://treendly.com/

3. Glimpse: Trend discovery platform for marketers and entrepreneurs. https://meetglimpse.com/

Market Research Tools

1. Exploding Topics: Trend discovery and analysis platform. https://explodingtopics.com/

2. TrendHunter: Trend spotting and innovation platform. https://www.trendhunter.com/

3. SparkToro: Audience research and intelligence platform. https://sparktoro.com/

Product Validation Tools

1. Validately: User research and usability testing platform. https://validately.com/

2. UserFeed: Customer feedback and product validation platform. https://userfeed.io/

3. PreLaunch Hero: Pre-launch marketing and validation platform for startups. https://prelaunchhero.com/

Pitch Deck Creation Tools

1. Beautiful.ai: AI-powered presentation design platform. https://www.beautiful.ai/

2. Gamma: AI-powered presentation and document creation tool. https://gamma.app/

3. Pitch.com: Collaborative presentation software. https://pitch.com/

Startup Tools

1. Stripe: Payment processing and financial services platform for businesses. https://stripe.com/

2. Product Hunt: Platform for launching and discovering new products. https://www.producthunt.com/

3. AngelList: Platform connecting startups with investors and job seekers. https://angel.co/

Rohan Paul

@rohanpaul_ai

Somebody uses an AI Bot to AUTOMATICALLY apply to 1000 JOBS in 24h and get 50 INTERVIEWS!

And the Bot's code is open-sourced in GitHub.

This tool automates your LinkedIn job search and application process.

- Scans for openings matching your criteria,

- Applies to jobs…

3,256

8:58 PM • Aug 27, 2024

Here's how this AI wingman works its magic:

1. You tell it what you're after - job titles, salary ranges, industries you prefer and those you refuse.

2. You feed it an exhaustive version of your career history.

3. It goes on a LinkedIn scavenger hunt, finding roles that match your wish list.

4. When it spots a winner, it whips up a custom resume and cover letter tailored to that specific role.

5. Then, it fills out the application and submits it complete with your tailored resume and cover letter.

6. And because it never sleeps, it keeps at it. Again and again and again.

Nothing New Under the Sun... Or Is There?

That's just what we're supposed to do anyway, right? You're not wrong. The gold standard of job hunting has always been:

• Customizing your resume for each role

• Highlighting the parts of your experience that make a persuasive argument for you as a candidate

• Making your resume into a highlight reel, not your life story

That's a ton of work. And even with all that effort, we all know how fickle the job market can be.

The Recruiter's Dilemma

Recruiters and HR folks are dealing with their own chaos:

• Sifting through mountains of resumes

• Trying to spot the needle in the haystack without picking a turd

• Racing against time to fill roles

Their filtering techniques have evolved from "Oh, fancy school? Big-name company? You're in!" to complex heuristics scoring applications on multiple dimensions. It's not perfect, but I get it. They aren't trying to find the best candidate. They just want to find an adequate candidate and avoid making a bad hire. You could be an absolute star but if there are already acceptable candidates in the pipeline when you submit your application, you missed it.

AI to the Rescue? (Maybe?)

The best case for an application is always getting personal intro or separating yourself from the pack in some other way. But when you can't, job hunters can now play this game at scale too. It's like a high-tech battle royale:

• Job hunter robots (these AI application tools) vs. recruiter robots (those scoring algorithms)

• Your meticulously crafted AI applications vs. their state-of-the-art filtering systems

• May the best algorithm win!

A Word of Caution

Now, I've heard about people getting temporary bans on LinkedIn for going a bit overboard with this tool. I still feel like I have to give it a try. For science. While working with AI, I've had moments of complete transcendence as well as utter stupidity. Where will this land? Oh, sweet baby Jesus, just please don't let it make something up.

The Real Takeaway

Whether you use this AI wizardry or not, here's the bigger picture:

1. Document every win, no matter how small.

2. We now have tools that can effectively summarize all the digital meeting notes, emails, and slack messages you are comfortable piping into it in order to providing fodder for review and promotion discussions.

3. When you're ready for that dream job, you'll have a treasure trove of accomplishments to draw from.

So, what do you think? I will definitely share updates if I find the nerve to give this thing a go.