Would you trust AI agents / assistants with autonomous tasks, or would you keep a close watch?
With AI agents and assistants are becoming more advanced, we're seeing them handle everything from scheduling meetings to managing entire workflows.
But here’s the big question — would you fully trust an AI to run tasks autonomously, or would you prefer keeping an eye on things just in case?
For example, would you let an AI handle customer support, process automation, or even financial decisions without constant supervision? Or do you feel there’s always a need for human oversight, no matter how good AI gets?
Three agents that stood out on PH to me:
ChatGPT Operator (not globally available yet)
Perplexity Assistant (Android Assistant)
Proxy 1.0 (Apparently better than OpenAI's Operator for specific use cases)
Have you tried any of the above or any other agents / operators?
@tanmay_jain_ I feel ChatGPT could lower its pricing, considering that power users often juggle multiple subscriptions rather than just sticking to one AI app. In this regard, Proxy clearly has the edge over OpenAI.
AI is getting smarter, but do we really trust it to run things unsupervised? 🤔
While AI agents like ChatGPT Operator, Perplexity Assistant, and Proxy 1.0 are impressive, I still find myself hesitant to give them full autonomy, especially in high-stakes areas like finance, legal matters, or customer interactions where nuance is key.
For repetitive, rule-based workflows? Sure. But AI still struggles with edge cases, ethical dilemmas, and unexpected failures. If something goes wrong, who’s responsible?
Curious—has anyone here fully handed off a task to an AI without checking its work? If so, what was the outcome?
I would rather keep it for my internal tasks. CRM is very specific case and in that scenario I would like to have AI as a co-worker.
I know what people are like in customer support and it is very difficult job with people (million questions, some of them yells at you, threatens etc...) This is the way I would like to supervise.
I let AI handle customer support, but I make sure to review interactions regularly.
We would look at using it for customer success as a co-worker too. It will get smarter and more reliable, so at this stage, I would tend to steer away from 'full trust' (finance, legal, even sales). AI is so helpful when used to complement a task. Complete automation is an option though; depends on your personality and business needs.
For routine stuff (scheduling meetings, drafting emails, handling FAQs) i prefer AI take over. bcz It saves time and streamlines work. But for high-stakes decisions always me who take decisions :)
I think it’s all about balance. AI is awesome for handling repetitive tasks, but for things like financial decisions or customer support, I’d still want to keep an eye on things. I haven’t tried Proxy 1.0, but it’d be interesting to see how it compares!
I'm cool with AI running automations but with finances or decisions, I'd still step in. Even automated systems need a quick check sometimes.
Even with AI advancements, human oversight remains crucial. Ultimately, accountability falls on humans when things go wrong. AI is excellent for automating repetitive tasks like emails and meetings, but for significant decision-making, human supervision is indispensable.
I've tried Proxy and Operator and so far found Proxy better performing and more likely to complete the task, it's good as well because its free!