I'm considering building an AI assistant that can handles research tasks by making calls and sending emails. It would be a virtual research assistant available 24/7.<p>Some potential uses:<p>- Call restaurants to confirm they don't use certain ingredients<p>- Contact experts or influencers in various fields<p>- Email vendors of a specific type for product specs<p>- Call vendors to get updated prices and availability<p>- Find 30 leads for customer development for exploring a new market segment<p>Example: "Find 5 restaurants within 10 miles that don't use these oils in their cooking."<p>AI: "Hi, I'm [AI name], an AI assistant for [user name]. I'm researching restaurants that don't use certain oils. Could you please confirm if you use these oils in your cooking? I'm compiling this information for my boss."<p>My experience: I used to have a research assistant and this type of outreach/call/email research was something they helped with. I don't work with them any more, but I miss getting their help on outreach research projects. I think an AI could many of these tasks in 20 minutes, running multiple queries in parallel.<p>Questions for HN:<p>1. What time-consuming research tasks do you do manually now?<p>2. What would make you trust an AI to handle these for you?<p>3. Have you used a VA/EA/PA before?<p>I think the main challenge is that I don't need to do this kind of research that often. I think certain kinds of users might need it a lot more frequently though.<p>I'm especially interested in hearing from founders, researchers, journalists, and other people who frequently need to gather specific information.
I compiled a list of things I find annoying or useless about my job that I want a scientist bot to do/fix for me in another thread. See below:<p>* Planning my own travel, purchasing all my own tickets and lodging and having no help to do any of that<p>* Sitting in meetings with colleagues who are explaining to me why it's ok that the interns aren't getting paid because they are going to give them visa gift cards instead<p>* Editorialmanager.com website<p>* Having no itemized or explicit way to examine the expenditure in my grants so that when I ask admin staff how much money there is for X they reply with "you have enough money" instead of the actual amount<p>* Researchgate website<p>* Colleagues who are rude and condescending instead of going to therapy to wonder why they are insecure<p>* Expense reimbursement instead of having a purchase card attached to my grants<p>* Convincing older colleagues to use "new" technology like slack or GitHub
No. None of the things you listed seem particularly time-consuming to me. And I prefer to ask questions so I can follow-up immediately with questions that arise.<p>Nothing in my experience with so-called AI or digital assistants tells me they can handle even simple tasks like you describe, without introducing more friction and uncertainty.<p>If a bot emailed or called me I would delete or hang up. If the person asking for help or information can’t bother to talk to me I won’t bother responding.
I am an academic researcher.<p>If I received an AI-generated email, I would permanently banish the sender to the deepest pit in my spam folder. Either that, or I'd have my own AI chatbot keep replying with confusing and vague questions until I'd exhausted the sender's patience.<p>The signal/noise ratio in this world is already bad enough. Please, please, please do not do anything of the sort you are describing.
To answer the question in your title, No. And if someone who did use it called me, I'd dig into my legal options to smack down that crap. We get enough robocalls and spam as it is, we don't need to help that industry get any worse.
I can't imagine the current state of AI supporting these types of interactions (or humans wanting to deal with AI on the other side), but would be happy to be proven wrong. I suggest focusing on this single use case of yours and applying it to your idea: "Find 30 leads for customer development for exploring a new market segment." Whether an MVP is successful or not will be a good indicator of feasibility.
I would definitely use this. I recently bought a hammer and I'm so tired having to call people, send emails to ask them if they have any nails that I can use the hammer on. Having a personal AI assistant do these calls and emails for me is going to be a god blessing. I hope I'll find some nails.
Didn't Google already try this and it didn't work? <a href="https://www.theverge.com/2019/5/9/18538194/google-duplex-ai-restaurants-experiences-review-robocalls" rel="nofollow">https://www.theverge.com/2019/5/9/18538194/google-duplex-ai-...</a>
As someone with my background (check profile), I would never.<p>Recommandation is something. Action on my behalf is something else.<p>Accountability and attribution matter more than what people think. When law enfoncement and "Justice" get involved, with unskilled people, when the stakes are prison time, a reader may appreciate this comment.
I would not use an AI to make calls or send emails for me. I wouldn't trust them to do the job properly, or to represent me properly. Even ignoring those trust issues, I wouldn't feel right forcing other people to deal with one.
> I'm especially interested in hearing from founders, researchers, journalists, and other people who frequently need to gather specific information.<p>I wonder if those types of people would even be willing to trust any responses?