Hacker Newsnew | past | comments | ask | show | jobs | submit | more anilgulecha's commentslogin

IMO, if this ends up occuring, it will follow how other practitioner roles have evolved. Take medicine, and doctors for eg: there's a high bar to reach to be able to do a specialist surgery, until then you hone up your skills and practice. Compensation wise it isn't lucrative from the get-go, but can get so once you reach the specialist level. At that point they are liable for the work done. Hence such roles are typically licensed (CAs, lawyers, etc).

So if I have to make a few 5 year predictions:

1. Key human engineer skills will be to take liabilty for the output produced by agents. You will be responsible for the signoff, and any good/bad that comes from it.

2. Some engineering roles/areas will become a "licensed" play - the way canada is for other engineering disciplines.

3. Compensation at the entry level will be lower, and the expected time to ramp up to productive level will be larger.

4. Careers will meaningfully start only at the senior level. At the junior level, your focus is to learn enough of the fundamentals, patterns and design principles so you reach the senior level and be a net positive in the team.


> At the junior level, your focus is to learn enough of the fundamentals, patterns and design principles so you reach the senior level and be a net positive in the team.

I suspect that juniors will not want to do this, because the end result of becoming a scenior is not lucrative enough given the pace of LLM advancement.


There’s a sweet spot right now to be in. Early enfough career to have gotten in the door, but young enfough to be mailable and open to new ways.


1. Already true, no company will make the AI agent liable for its output, it’s always the programmer

2. Unlikely, as most software won’t result in death/injury… whereas a structural engineering project is much more life threatening.

3. I actually think entry level engineers will be expected to ramp up to productive levels much much quicker due to the help of AI

4. Already true


Had to laugh at #4… that’s where I thought we were now.


Canada?? They can’t build a subway station in 5 years nevermind restructure a massive job sector like this lmao


You're being downvoted but you're actually spot on

Calgary was supposed to have a new train line, planning has been in motion for years. Back in 2019 when I bought my house, the new train was supposed to open in 2025. As far as I know not a single piece of track has been placed yet. So... Yes


Yes this is status quo in Toronto but it’s a Canada-wide corruption issue.

https://www.cbc.ca/news/canada/toronto/eglinton-crosstown-op...


Hey, they knocked down Eau Claire. So.. progress?


Are you sure? Looking forward - AI is going to be so pervasively used, that understanding what information is to be input will be a general skill. What we've been calling "prompt engineering" - the better ones were actually doing context engineering.


If you're doing context engineering, you're writing an agent. It's mostly not the kind of stuff you can do from a web chat textarea.


Very interesting approach. Why a browser, and not a fantastic chrome extension? Grouping tabs, summarizing, even taking open ended actions, seem very doable with permissions extensions have..

edit: Just read about the accessibility thing, but that's thin. Is there any usecase in the future that a browser can, but an extension can't?


> Is there any usecase in the future that a browser can, but an extension can't?

The only reason to use a browser over a chrome extension is to bypass security features, for example, trusted events. If a user wants the browser window to go to full screen or play a video, a physical mouse click or key press is required. Moreover, some websites do not want to be automated like ChatGPT web console and Chase.com which checks if the event was a trusted event before accepting a button click or key press. This means that a Chrome extension can not automate voice commands inferred with audio to text. However, to get a trusted event only requires the user to press a button, any button, so message or dialog prompt that says, "Press to go full screen," is all that is required. This can be down with a remote bluetooth keyboard also.

The way I see it, these limitations are in place for very, very good reasons and should not be bypassed. Moreover, there are much larger security issues using a agentic browser which is sending entire contents of a bank website or health records in a hospital patient portal to a third party server. It is possible to run OpenAI's whisper on webgpu on a Macbook Pro M3 but most text generation models over 300M will cause it to heat up enough to cook a steak. There are even bigger issues with potential prompt injection attacks from third party websites that know agentic browsers are visiting their sites.

The first step in mitigating these security vulnerabilities is preventing the automation from doing anything a Chrome extension can't already do. The second is blacklisting or opt in only allowing the agents to read and especially to write (fill in form is a write) any webpage without explicit permission. I've started to use VSCode's copilot for command line action and it works with permissions the same way such as only session only access.

I've already solved a lot of the problems associated with using a Chrome extension for agentic browser automation. I really would like to be having this conversation with people.

EDIT: I forgot the most important part. There are 3,500,000,000 Chrome users on Earth. Getting them to install a Chrome extension is much, much easier than getting them to install a new browser.


It sounds like something that needs to be dealt with in Chromium rather than forked. I am sure lots of developers want such functionality, if it is missing. I found:

https://developer.chrome.com/docs/extensions/ai

Don't any of these fit the bill? Are they Gemini-locked and you want something else? I am not familiar with the Chrome API, so pardon my ignorance.


Yeah accessibility is one such usecase, but in future we have few other ideaswhere having a fork makes it lot easier. Few ideas:

- Ship a small LLM along with browser - MCP store built in


My personal insight on this: toy projects are fun not because they're small in scope, but because there's no downside to failure.

Production/professional software has significant downsides when they fail (customer unhappiness, professional performance and incentives, etc).

It's also why toy projects open up very innovative pathways. You tend to not be conservative with them, and once in a while something amazing ends up working. Bringing this non-conservativeness to professional software is a skill once developed makes your growth shoot up.


I'd made a prediction/bet a month ago, predicting 6 months to a full 90 minute movie by someone sitting on their computer. [0]

The pace is so crazy that was an over estimation! I'll probably get done in 2. Wild times.

0: https://www.linkedin.com/feed/update/urn:li:activity:7317975...


It's doable now. Someone just needs to do it. With voice now it's completely doable. Just throw it all together add some effects and you've got a great movie... In theory


It's not a theory, at Cannes a feature movie has premiered that is generated entirely by AI. Made in Spain.


It's a great movie in theory. Idk how good the movie you mentioned is


can you share a link/details, please?



There's still a lot of work to be done. It's good at making short individual scenes but when you start trying to string them together the wheels start to come off a lot. This [0] pretty basic police raid leads to shootout video for example turns to mush pretty quick because even in the initial car ride the interior of the car's size and shape warps pretty drastically.

Feels like there's going to be a dichotomy where the individual visuals look pretty good taken by themselves but the story told by those shots will still be mushy AI slop for a while. I've seen this kind of mushy consistency hold up over the generations so far, it seems very difficult to remove becasue it relies on more context than just previous images and text descriptions to manage.

[0] https://www.reddit.com/r/ChatGPT/comments/1kru6jb/this_video...


Also, would it bring more to table on the web compared to svgedit?

https://svgedit.netlify.app/editor/index.html

https://github.com/SVG-Edit/svgedit


Chrome Devtools (and firebug) are classic, well thought out dense interfaces. so are VSCode, Jetbrains IDEs.


Key truth: Parents, on average, request for this. They hold the school/university (not their wards) accountable for anything that happens to their wards.

There are forward thinking universities/academicians as well. Here's an article by an academician: https://www.deccanherald.com/opinion/india-s-universities-li...

(We run 25+ UG programs in India. If someone would like to engage deeper on this with ideas on how we can improve, pls reach out)


Thanks for pointing this out. Indeed, there are Indian universities (including the one I went to) which do not do this.


Exactly how EC2 is different from Lambda from a user's perspective.


Under-rated, but probably the best serverless offering: Cloud run on GCP. Pay like for a VM, but only for the time you're getting/serving requests.

(IMO, if it can get a fly.io like command line experience, it will thrive more.)


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

Search: