Brian Amerige @brianamerige

@tjparker Beautiful. I did a similar thing. Had a 250yo blue oak fall in my yard. I was heartbroken. Used to be able to see it from almost every window in the house. Irreplaceable, but I wanted to turn into something special. Transplanted a 35yo, 30k lb live oak. Such lovely memories.

keyboard_arrow_left Previous keyboard_arrow_right Next

@WasatchSnow Not sure about measurements but it’s also been snowing all day in Canyons at PCMR. Still is! Been a great day.

Me too. It was the first time I voted for a major candidate—and not because I like Harris even a little bit. I literally moved states to free myself from Newsom (and his ilk). Was a surreal experience for me, but I can’t look past the GOP aligning behind someone who is explicitly and empirically an enemy of the Constitution and truth.

I rarely tweet, but this is one of the most obviously ridiculous takes I’ve seen from you. Implying that a state’s GDP is a function of its governance is just silly. Productivity is affected by governance, but it isn’t determined by it. There are many historical factors that have made Silicon Valley what it is, but generally good governance is obviously not one of them. California has high GDP _in spite_ of its almost unbelievably bad governance.

@awilkinson Happy to forward to one of my friends still there. They have a dedicated channel internally for things like this.

It's baffling this needs to be said, but after reviewing 600+ software engineering job applications, it clearly does: If you're applying for jobs and using ChatGPT to help answer questions like "why are you interested in this job?": just… don't! You are wasting your time and will be immediately (if not automatically) rejected. You *probably* think you are doing yourself a favor by using "AI" to help you write "custom" applications to hundreds of companies. No. You are wrong. Your LLM-generated answers are: (a) obviously written by an LLM, and identifiable as such in less than 3 seconds (b) nearly identical to everyone else using an LLM (c) generic, unconvincing, uninspired, and bad I get that the job market is crazy competitive, but I swear you are better off with 50 actually-thoughtful applications than 500 made possible by this garbage.

Thoughtful is hiring eng #2. We've kept the team tiny to date, and I'm really glad we did. It gave @AlexEpstein and I the space to iterate quietly and make sure we had something that not only has serious engagement+retention, but actually works to transform wasted downtime into time spent learning and growing. Now, we're gearing up to focus on growth and monetization. We need another engineer to take on significant areas of ownership across our native iOS app and backend. Details below👇 jobs.ashbyhq.com/thoughtful

x.com/i/article/1811…

@DanielNolin3 @AlexEpstein 🙌

Bad 1:1s end up as status reports. Good 1:1s: * are driven by, and for, the report, not the manager * should mostly be about things that can’t or shouldn’t be said in front of everyone else * should be focused on the big picture from the perspective of your report—like what they’re excited about (or struggling with), and how their work fits into your team and company’s mission. * have as a byproduct of 1:1s a map of how to get where they want to go. Managers grow a team’s capability by growing ICs’ capabilities and project/goal alignment. * another byproduct is trust. A great team doesn’t just do their jobs. They know each other, trust each other, and like each other. People need time together to foster that. All of this kind of thinking is crucial to great work but requires prompting in most people. It’s a huge departure from the tactical tunnel vision most people have when executing with intensity.

@thogge 💯 From tonight’s walk with my dog. Unreal!

keyboard_arrow_left Previous keyboard_arrow_right Next

It’s a spicy take but today’s AI may be useful, powerful, etc., but it’s not intelligent by any meaningful definition. And the pursuit of AGI via LLMs is a non-starter that is, unfortunately given the level of investment today, a waste of effort 😬

Tsarathustra @tsarnick

2 years ago

François Chollet says AI language models have a very low level of intelligence, as measured by the ARC challenge

28 32 195 22K 88

I built and sold Frame.io for $1.3 billion to Adobe, and along the way, I received plenty of advice that turned out to be entirely off the mark. Here's a look at some of the most glaring misdirections: 1. "As CEO, focus on executive business tasks." This couldn't be further from the truth. The core purpose of any company is the product it offers. As CEO, your primary role is to ensure that this product is the most valuable it can be for your customers. It's not just about overseeing; it's about being integrally involved in delivering quality. 2. "To scale, you need to decentralize decision-making." Actually, the opposite is often more effective. As organizations grow more complex, centralized decision-making can slice through red tape and foster quicker action. Even in environments that champion decentralized processes, top-down leadership often proves to be the fastest route. 3. "CEOs shouldn't get caught up in the details." This is a myth. Consider leaders like Steve Jobs, Elon Musk, Mark Zuckerberg, Brian Chesky, and the Collison brothers — all are/were deeply involved in the intricacies of their businesses. Emulating their approach is not misguided; it's a blueprint for hands-on leadership. 4. "10X engineers are a myth, and believing in them is harmful." Quite the contrary. In many teams, a small fraction of the workforce often generates the majority of outcomes. Recognizing and nurturing high-performing individuals isn't just realistic—it's crucial. 5. “You must validate every decision with customer research.” Contrary to popular belief, not all successful decisions stem from extensive customer research. Many of my pivotal choices were guided by my own instincts and preferences for the product. Trusting your vision and building a product that you would use yourself can lead to creating something that resonates deeply with your customers. Personal intuition can be a potent tool in product development.

@garrytan Do you think it's becoming a lost art for consumer-focused founders too (i.e., those building things normal people actually want)? My guess: selection bias at play here—AI-gravy train founders are unlikely to care about user experience.

@mbrendan1 We’ve been building a new kind of “go to” app that transforms toxic, usually-wasted social media downtime into time spent going deeper on your interests, learning, and growing. Very much carefully optimized for human flourishing, not engagement. thoughtful.community/brian

@mbateman A few years in with @AlexEpstein…I agree. I’m as philosophical as founding teams usually come, but Alex’s standards for clarity slow me down at the right moments for the right reasons.

HTML and PHP at 11, making websites about the glitches in Xbox’s Midtown Madness 3 😆 Mindstorms at 12 as well—part of the robotics “competition” I had at school. And then started on Objective C at 13 to build a Mac App so members of the forum I hosted could create their own websites.

Also, hi @MikeIsaac. Good to meet you on Twitter. Hope you have been well since the FB political diversity days :)

I'm a founder also building in the "tech that helps us focus more on what's important in life" space. So I not only like builders, I get this space. Ai Pin is more than merely "not there yet." It fundamentally misattributes "toxic screen time" to "backlit screens." Okay, fair enough. Those mistakes happen all the time. No biggie. Except what @Humane actually did was produce irritatingly widespread, vacuous hype for years before finally(!) releasing a product that solves a problem people don't have. I hate mobs. And I won't defend personal attacks. But highly critical reviews? Appropriate, IMO…

rat king 🐀 @MikeIsaac

2 years ago

in tech land theres a whole furor over this AI-powered pin called humane that was universally panned by reviewers for not functioning founder types are getting upset that folks aren't supporting "builders" i think the mistake was the framing and high profile launch

46 13 494 102K 30