Comment by NichoPaolucci
1 day ago
What exactly do we want the general public to "do" with these models? If the have no use for agentic LLMs, they obviously aren't going to be interested in the capability of those agents.
I see a lot of benefit to spreading information around the tech, but it's still in its infancy. Most people weren't on the internet when that was getting up to speed.
Meemaw and pops have their systems that they work within, and they just got familiar with the damn smartphone. Though it might help them, they probably do not want AI to organize their old photos or "clean up" their computer. Saving 5 minutes just isn't important to them.
As for professionals, yeah - companies are going to adopt these tools over the coming years. It's the "norm" in my circle, but lots of jobs just... don't use technology like that - there has ALWAYS been a major gap between people who are deep in tech and people who are not.
Personally, I think the worst possible use case is the one that's being pushed for; personalised agents.
I think a much better use case is one where it's not visible to the user at all, like spam filtering or helping an accountant do their monthly run in a few hours rather than a day. To me, that's likely a much better use of AI than trying to shove a chat interface or agent harness into an existing app.
That being said though, I did see a startup in my area basically offering a Resend/Mailchimp competitor as a Claude plugin, so maybe it'll just be integrations to their existing chat interface.
I've always struggled with doing basic life stuff. Paying bills on time. Remembering birthdays and to buy presents. Doctor appointments. Setting up play dates for the kids. Getting my teeth cleaned regularly. Booking in vaccines for the kids. Renewing passports. Booking flights. Etc. It feels like so much of my life is consumed by this useless administrative bullshit. I LONG for personal agents which take away all this friction from my life so I can focus on the things I love: my family and my hobbies. A personal agent will revolutionise my life in a way no other technology has.
At the expense of your privacy, potentially acting irrationally on your behalf, with the inherent weakness of being able to be prompt injected. All of that is not even to mention the grey area of who is responsible when your agent breaks the law.
These aren't hypothetical, and in the hands of non-technical users let alone technical users, they have are far too unpredictable.
I don't think 99% of people have any application for AI as it is today short of personal assistants. But, if we can have recursively improving systems for that 1% of people that are responsible for scientific and technological innovations, that's going to be transformative for everyone.
I'll assert that 99% of people don't have any application for personal assistants, outside of work (perhaps). Most of us aren't so buried in personal obligations that the value of a half-assed "assistant" (which might violate our privacy) is worth the tradeoff in complexity or security.
The pitches for Dots, Grok Bot, etc., have been spectacularly uncompelling for me as a civilian. I don't need or want a bot to book my flights or read my inbox. I'm not so busy that these things have a value high enough to be worth the effort.
As a worker or an entrepreneur...maybe. Depends on the task, and how many other people were flinging slop at me on a daily basis. I could see the value of a calendar bot, or an expense reporting bot or a bullshit paperwork bot or a slop-return (slop jai alai?) bot, but there's no way in the world I'd give a bot access to the information necessary to, say, do my personal taxes.
Well, you can't run a personal AI assistant with 640k of memory.
1 reply →
i'd argue any parent, especially of a young child, could use 20 personal assistants lol. So that's just false. There's SO many appointments to keep track of in baby's first year of life, and that's if you have a healthy baby.
And once they start school there's an endless amount of deadlines and "reminders". A personal assistant that can automatically keep your reminders + calendar up to date based on your inbox would be very nice.
It gets even more fun if you have an aging parent at the same time as you have young kids. Then you get double the appointments to scheduled, coordinate and keep track of.
So when you say people don't need a personal assistant, i feel like you're living on a different planet. I suppose when i was single things were more "simple". Your mileage may vary of course...
5 replies →
I have been kind of disappointed by AI science-wise. I was anticipating more advancement. Experiments do take time, so I suppose they could be pending.
AI really struggles with "collision with reality". Computers suck at modeling the real world in a detailed enough way to make scientific progress, aside from small niches like drug discovery where you have a very well scoped problem. If any part of the try something -> test it -> look at the results loop requires hitting the real world, AI is going to struggle.
On the other hand, the areas where it doesn't need to hit the real world, progress is absolutely flying. I would not be surprised if AI is superhuman in math and physics in a year, considering we went from homework helper to IMO Gold to Millennium problem solving in a couple years time. I think we'll start to see more meaningful scientific progress in maybe 2-3 years when scientists have access to an oracle that can formulate theories and direct real-world research better than a human can.
1 reply →
My hypothesis is that the help/support buttons in nearly all apps will be to an agent in the future. Like now you would type "how do i do X" in the excel search bar, in the future it will be "do x". Users in 10 years may not even know how to use the menu/mouse functionality for apps.
Why would you think this? Giving orders to software is such an ineffective, time consuming, error prone way to interact compared to using the interface yourself. That sounds delusional, and would be such a waste of resource, horrendous engineering and user experience
If you know how to accomplish it using the interface, yes
My grandmother keeps a little notebook with detailed instruction on how to do things such as sending a message on facebook or pay a bill using netbank.
For a lot of things, the choice isn't going to be doing it herself or giving orders to the computer. Its going to be not doing it or giving orders to the computer.
Even if it will look very inefficient to people such as you or me
I didn't think it was that controversial actually. I don't bother navigating man pages or help FAQs anymore, i just ask the agent how do i do X step by step. Opening this up as part of the app to regular non-engineering users seems relatively easy to implement, but since it would have app permissions you would skip the "give me the steps" and just say do it.
It is jobs to be done theory on steroids. If someone wants a drill, give them a hole, this kind of logic. Let users literally click “make a hole” button that would instantly make a hole in the wall, and you’ve just disrupted the whole drills market.
Or like in Harry Porter, where you don’t need cooks - food just appears from thin air and floats straight to the tables.
The problem is, jobs to be done theory is not new, and product teams that build Excel and SAP know this stuff. The reason why nobody has yet made an accounting software that is just one magic button is in the amount of user decisions that are necessary to make a “proper hole”.
This is where analogy breaks, because you are not helping to drill a hole anymore, you are helping to design a hole. And to design something you need visibility into what parameters you control, what has just changed with the latest tweak, where rough patches are that require cleanup and design rework.
In programming it just so happened that your code structure and your IDE were also your design UIs. In other areas, like Accounting, your design UI today is Excel and SAP.
Arguably a chat interface can also be a standalone design tool, just not a very user friendly one, like trying to pilot a plane through popup notifications. It works for areas where we fully give control over design to LLMs themselves, but step up into higher level designer shoes and you need proper UI/UX again, chat becomes suffocating.
But would you not agree that writing less code by hand and even looking less at code at all becomes more practical with each new model? If we extend this trajectory, we should see that the time interacting directly with code becomes less and less.
I don't think his point is that everyone should be using AI, it's that if you haven't used the agentic tools with a frontier model in the past month or so, you have no frame of reference for the current level of capabilities. Most people only use the chat interface with the free-tier models and based on that experience they're making strong conclusions about where AI is headed (it's still just fancy autocomplete, the whole thing is a scam or a bubble, etc.).
> if you haven't used the agentic tools with a frontier model in the past month or so
We keep hearing versions of this same drivel since late 2025. This time it's different! Now it's really super duper amazingly awesome powerful and if you don't believe it, just trust me bro.
We keep hearing versions of your comment, despite recent breakthroughs in hundreds of open problems that mathematicians have been trying to solve for the past century. Skepticism is good up to a point, but it needs to be amenable to new evidence. You need to keep your mental model up to date with reality.
Don't trust me, experimenting is cheap; get a $20 subscription at Anthropic or OpenAI, download Codex or Claude Code, and try the heaviest model on a project that you estimate will take you around a week to complete without AI. I was a skeptic until I did exactly that and it shifted my perspective by quite a lot.
I mean, the parent poster is not wrong though. It’s crazy the level of stuff you can offload to your agents. People who haven’t used these tools in the last short while are operating on old, outdated frames of reference.
I’m just about to tell opus or sol to go make a custom watch holder that fits kiddo’s bed frame. I’ll give it the measurements of the bed frame. It already knows the dimensions of my printer. I’m prepared for it to absolutely fail at the task, but honestly I wouldn’t be surprised if it came up with something I can send to my printer and just work. Part of me wonders if I could instruct it to simply generate the raw gcode and skip the entire stl -> slicer pipeline. Probably would blow its context window though.
If you haven’t used these tools, you really don’t understand their capabilities. It’s pretty crazy stuff.