@buildwithpb My own products grow through organic traffic mainly. I barely run ads for myself, but I managed ad campaigns for my clients who are seeking product-market fit. Once they validate their business concept, they ramp up advertising spending to boost overall revenue.
Running ads is an efficient way to quickly verify product-market fit, as organic traffic is difficult to attribute.
Even ChatGPT runs ads on Meta, TikTok, and YouTube.
I submitted both the iOS and macOS builds for App Store review together . The macOS version has been approved, while the iOS version is still pending review.
@solopribuilds The issue is that achieving better results on this platform would require our full time and attention. Focusing exclusively on one channel carries the risk that it may not be the most efficient allocation of our resources.
Back in January 2025, I still had to provide specific technical instructions to ensure that AI-driven bug fixes wouldnโt compromise the entire system.
Now, however, most models can autonomously review and resolve bugs on their own. AI is evolving at such an incredible pace that I can genuinely envision a future where a single document is all that is needed to build an entire product.
I'm officially done reading AI-generated code.
It's been two weeks since I looked at any of it.
I think the IDE is officially on its way to the graveyard. The job is no longer about "writing code," so we need new tools that better reflect this new reality.
While reviewing the code, I realized my only complaints were stylistic, and I wasn't finding any obvious bugs anymore.
The more code I generated, the harder it became to keep track of every line. I found that my time is better spent designing ways to verify that the overall system works than looking at the code.
State-of-the-art coding agents are better at writing code than I'd ever be, and I'm going to stop pretending otherwise.
I still think these coding agents can't go too far without an experienced human guiding them, but we're past the point where we need to check every line of code.
Based on my experience building products with these models, Grok 4.5 produces high-quality code for basic tasks.
I would recommend using GPT-5.6 or Opus 5 for the planning phase, and then delegating the execution to Composer 2.5 and Grok 4.5. This workflow could save a substantial number of tokens.
@gregisenberg An effective text message is a powerful way to convey product values. If the messaging doesnโt resonate with people, they wonโt be interested in the product, no matter how polished or beautiful the interface may be.
@Mayhem4Markets Six months ago, I built an AI-powered reply bot, but I soon realized that letting AI speak or represent me is risky, and I donโt want to be replaced by a soulless robot. AI lacks a true perspective and simply synthesizes existing content.
Iโve spent thousands of dollars on various user research platforms, but honestly, the best approach is always to reach out to your current users. You can simply have them book a time slot directly on your @Calendly.
If thatโs not an option, @userinterviews is your next best bet. I prefer it over @Prolific because it allows participants to share their screens, so you can watch how they actually use your product while discussing it with them in real time.
@boardyai expressd interest in my project and suggested we jump on a call. This AI robot felt so real in todayโs Google Meet that I am considering integrating a similar bot into my own products as a customer service solution.
We are launching new campaigns in Boston this fall semester. If you are a university student looking to earn commission by offering your skills or services, we invite you to join our ambassador program.
During the spring semester, we successfully onboarded students from Boston University and Northeastern University to buy and sell services on our platform.
This fall, we are expanding our reach to include Harvard, MIT, Boston College, and more.
@sama Would OpenAI ever consider open-sourcing a model comparable to DeepSeek V4 or Kimi k1.5? Could ChatGPT bring back the original GPT-4o? For writing tasks, I still believe it is the best model, as it strikes a perfect balance between a human-like tone and logical reasoning.
Business owners, investors, influencers tend to prioritize different metrics. I generally avoid following those who donโt actually build anything themselves but spend their time bragging about the latest tools or how to use them. While they may be skilled at social media marketing, they lack the practical experience required to build a product or scale a business.