In my own experience, which I'd recommend you disregard if it isn't useful for you.
Try lots of things, see if something sticks, it might take years, or decades, but the variety of trying many things is interesting in itself.
Try to connect the dots in your own life, for example what were you like as a kid, what interested you, has any of this followed into your adult life, what about looking forwards e.g. where do you want to be 30 years from now and can your current or new interests get you there.
Having kids will certainly give a person purpose and eat up all of that free time if that's in the plan for you.
What a range of really interesting responses in this thread, thanks for taking the time to post. In 10 years I'll be approaching a milestone age (like I am now). I hope to still be an active father for my kids, to have enough money and investments to feel financially comfortable, to be pursuing more of my interests than work.
I got a Masters in Software Development, part time evenings, studied over 3 years. I don't think it's strictly necessary, but in my experience, it doesn't harm your prospects. I think it's nice to have the combination of experience and education, personally. Certainly don't underestimate the commitment if you study in your spare time. Also, you may enjoy the course, work on new things, meet new people - all can be valuable in their own right.
Congratulations on launching, great idea. Can I ask you about AWS lambda for the email sending. Did you integrate with a third party email service, or node mailer, or just a Gmail (I believe they get throttled at 500 emails per day). Or does AWS send the emails? I'm working on a project that needs emails to go automatically and I'm not sure on the best implementation yet, I'm leaning towards the Gmail option but don't want to hit the limit. Thanks
As an alternative you could appoint a software services business to build it for you, to your specifications, then they step back responsibly handing it off to your business. It's more expensive but they will be off your books in 6 to 12 months unlike if you hired your own team. You'd hire a small team to maintain the system, rather than a large one to build it.
I wonder if there is a technological way to reduce this number, and also if the number was reduced, how this might impact the environment through hundreds of millions more birds living.
I came to the conclusion, personally, that a non technical co-founder should spend the time to become technical. I'm sure it's not needed in all cases, but for me, it was the right path.
If you go to the OpenAI developer forum, and search for this thread, there are lots of academic papers "Foundational must read GPT/LLM papers". I just graduated from a masters a few days ago, so can appreciate the hard work involved. Good luck, it'll be worth it!
Following up on my own question. I re-read the github, so I can clarify my question better. So the AI agent responses are saved to a database where the human sme can classify responses as good/bad right? Do you intend for the result of this analysis to retrain the AI agent, or is it purely to get a baseline on the as-is AI agent quality?
Hey team, nice work. Can you help me understand this better. How does the process work in terms of the human agent evaluations? Is it real time so that the right (maybe a better word is best) answers go to users as they are needed, or is it done asynchronously/batch style so that the humans are training models to be better? Once the best answers are selected, is it fed back into an LLM / AI agent model? Thanks
Personal home pages for individuals and families, they were always warm and friendly. Also, internet cafes, the first time I used the internet I went to one and looked up the X-Files homepage. I also remember chat rooms where you could use a telephone handset to talk to people.
To help me with something similar, I score my ideas against a number of measures and questions. It's helped me choose something to build over the last 6 months or so.
They are;
### Idea Validation
- Is this - Unworkable (consequences for the user are serious if not addressed)
- Is this - Unavoidable (something the user needs to do)
- Is this - Urgent (and needs to do now, number 1 thing)
- Is this - Underserved (MVS isn't served)
- What is the Minimal Viable Segment
- Is this - Blatant and Critical?
- Is this - Latent and Aspirational?
- What are the - Dependencies
- What are the - External Factors
- Could there be a - Backlash
- Is the timing good
- Is this - Discontinuous
- Is this - Disruptive
- Is this - Defensible
- Describe the - Before / After
- Is this - Gain / Pain > 10
- How big could it be 1 - 10
- Founder / Market fit (am I an expert?) 1 - 10
- How sure are you that this is solving an acute problem & do I have personal experience 1 - 10
- Have a new insight (maybe counterintuitive insight) 1 - 10
- Has this only recently become possible?
- Are there network effects
- How frequent a problem is this
- Is this a 0 to 1 idea
- Can this be a monopoly
- Can it make money
- Could I work on this for 10+ years
- Could I build this
### Value Prop Statement
- This is for (target customer)
- who is dissatisfied with (existing solution)
- due to (key unmet needs)
- We offer (venture name & product category)
- That provides (benefits of solution - not better, faster, cheaper)
Many larger firms who use agencies will go through a competitive tender process, and for all who are technically compliant, they choose the cheapest. It's worth considering how to navigate that if you want to be high end (which I'm presuming means expensive). Also, I'd challenge 95% of firms are terrible, what have you based that on?
Can recommend. This is how I did it over the last year. VSCode, Python, Flask, ChatGPT, and pick something challenging to build. Keep going for a year.
Thanks. Can you share a sample of the questions you used? Were you comparing the LLMs to answers you knew already, or comparing the new answers which were output by the LLMs