Hacker News new | past | comments | ask | show | jobs | submit | isaacisaac's comments login

I've been working on Temple Tools[1] (CRM for Synagogues) during evenings and weekends while maintaining my day job with no drops in productivity or outcomes (in fact I've performed better) for about 9 months and have gotten a good amount of traction. I was really afraid that it was going to take over my life but I'm very happy with how straightforward things have been. Building a CRM solo has been quite a lift but it's gone quite well.

Here are my tips:

• Get your first customer as early as possible. I know this isn't "new" advice but feedback from one customer was enough to keep my busy patching holes in the product for a good 2 months straight. If I waited and got my first 10 customers all at once I would have drowned

• Separate functions into days of the week as much as possible! I've found this to be critical for not burning out and maintaining high productivity over extended periods (9 months doesn't sound like much, but it's a long time to put in essentially an extra work week). For me Mon is Marketing, Tue-Fri is coding, Sat is marketing and non-urgent customer support, and Sun is coding

• Don't stop exercising. Exercise MORE. I'm a night owl but I transitioned to the "at the gym at 5am" guy, lunchtime run, and Jiujitsu in evenings; and I ended up somehow spending more time plugging away at the keyboard. The mantra "the way you do anything is the way you do everything" really comes to mind.

• Put dates on your calendar for meta-tasks. I have a recurring bi-weekly task to update my finances, a task to do keyword analysis, etc etc. It's too easy for these to slide otherwise

• Creative uses of ChatGPT! I'm using Nylas' free tier and the ChatGPT API to automate triaging my own customer support emails, and texting myself if it's urgent. This lets me have near-zero interruptions during my day job from my side hustle.

• More ChatGPT. If you don't know how to do something, talk to ChatGPT interactively to get "good enough" at it quickly. This could be anything from finance to SEO to cold outbound. You still have to think critically enough to ask good questions like "what am I missing here" or "what's going to bite me in the ass". You absolutely need a paid subscription (I use both ChatGPT4 and Claude)

• Be your own project manager. Do NOT wing it. Sit down every once in a while and list off everything you think you have to do, break it down, triage it, prioritize it, and start burning it down. This includes both product-level tasks and business administration and marketing. This lowers cognitive load so much that you'll find you can actually take work off your mind when you're not working

• Identify opportunities for crunch times and adjust routine. I've rented an office on three separate months while working on this business. It was a fantastic way to break up the routine and get a huge amount of productivity for a short time

1 - https://temple-tools.com


Great advice!


People skills are transferrable to prompt engineering


For example, my coworkers have also been instructed to never talk to me except via code comments.

Come to think of that, HR keeps trying to contact me about something I assume is related, but if they want me to read whatever they're trying to say, it should be in a comment on a pull request.


I've heard stories about people putting this garbage in their systems with prompts that say "pretty please format your answer like valid json".


I guarantee LLMs are a big piece of it. Not necessarily ChatGPT-like interfaces but instead dynamic interfaces/behavior that intuit what you want the device to do, especially on mobile where the user <> device interaction bandwidth is quite constrained.


This is pretty much confirmed: https://www.bloomberg.com/news/newsletters/2023-10-22/what-i...

>the company built its own large language model called Ajax and rolled out an internal chatbot dubbed “Apple GPT” to test out the functionality. The critical next step is determining if the technology is up to snuff with the competition and how Apple will actually apply it to its products.

>Giannandrea is overseeing development of the underlying technology for a new AI system, and his team is revamping Siri in a way that will deeply implement it. This smarter version of Siri could be ready as soon as next year, but there are still concerns about the technology and it may take longer for Apple’s AI features to spread across its product line.

I wonder, will they launch this as "Siri, but actually useful" or will the Siri brand be replaced altogether.


If this next-gen LLM-based Siri is built to take advantage of local documents/data with all processing happening on-device, I think it could be more useful than incumbent LLMs in some ways even if it’s technically not as smart simply because it has far more relevant context by default. It being local-first could also help assuage privacy concerns and make users less reluctant to use it.


I'd be happy if they used an LLM interpret the command so Siri could better understand what to do without specific wording or requiring multiple separate commands.


I work at Nylas (Gmail/MS/IMAP email & calendar connectivity APIs) on the product side and I really wish something like this received adoption.

Between the insufficiencies of IMAP for server-side use and the disparities between the "winners" of the space, the lowest common denominator of what's supported for email is abysmally small.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: