Hacker Newsnew | past | comments | ask | show | jobs | submit | rukuu001's commentslogin

This is the answer, and this strategy can be used on lots of otherwise unsafe activities - put a tool between the LLM and the service you want to use, and bake the guardrails into the tool (or make them configurable)

Well, be careful. You mmight think that a restricted shell is the answer, but restricted shells are still too difficult to constrain. But if you over-constrain the tools then the LLMs won't be that useful. Whatever middle ground you find may well have injection vulnerabilities if you're not careful.

It shaves yaks for me. I appreciate that.

Hey man, just wanted to say thanks for Selenium - it was a game changer and had a big impact on my professional life.

I’m interested in checking out Vibium - I’ve been a reluctant adopter of Playwright and hopeful for a new approach.


playwright got a lot of things right. one of the big ones was a fast websockets+json way to drive the browser. (vibium is using the w3c standard equivalent - webdriver bidi). but they also raised the bar on usability and developer experience. i hope to get to the level of "click, click, awesome" out-of-the-box experience that playwright did so well.


If you're already using Playwright, I'd love for you to give Stagehand a try (https://github.com/browserbase/stagehand) - it has compatible-ish API, but built for automation, not testing.


Hmmm so it is basically using https://www.director.ai for the AI natural language stuff right?


Director.ai uses Stagehand, which is made by Browserbase.


stagehand is good stuff!


> I’ve been a reluctant adopter of Playwright and hopeful for a new approach.

Out of curiosity, why?

Personally, I'm a massive lover of playwright. Flakiness has been so much lower for us.


I think the future is mobile when it comes AI agents, If you are also looking for a new approach, I would suggest DeepWalker. It is mobile first automation, currently works on android.

[0]: https://deepwalker.xyz


Think of highly competitive environments where looking foolish can be weaponised against you. They definitely exist here (my experience in UK and Australia)


IBM Aus discouraged it, Accenture, Concentrix, EY, CommBank, ANZ, and others all encouraged it, for myself.

I wouldn't say discouraging it will be the norm across most places in Australia.


Somewhat, but not exactly the reverse, is Tall poppy syndrome:

https://en.wikipedia.org/wiki/Tall_poppy_syndrome

I had read it about AU and JP and had read about the Jante thing, but the article says it is there in some other countries too. Probably exists everywhere in some form.

I wonder if people here have experienced anything like that. My guess is yes.


That, I have certainly seen.

But that is about attacking success stories, not about attacking you for not knowing something. I know you said reverse, just spelling out they're different.

Win an award? Get a callout for effort? Rest of the office will probably dunk on you. Varying in scale from a day or two of jokes to career ruining.

But... Ask the meaning of an acronym in a meeting, or say you don't know how to do something, and you'll probably just be given a name to ask.


>I know you said reverse, just spelling out they're different.

Hey. Sup. :)

I didn't say reverse.

I said:

>Somewhat, but not exactly the reverse, is Tall poppy syndrome.

Next time, and forever after, check points better before commenting, including the point that is right above the comment that you replied to, aka mine.

Pinky promise? :)

Welcome to the AI (r)age, where people comment without thinking or reading.

Oops, mea culpa.

Entschulding!

It was happening much before that, and not only on HN.

In fact, it has been happening forever, and anyone who doesn't know it is a rotten egg.

<Jumps into the pool ahead of others. (last one in, etc.)>


Well, this is some fun irony. An attacking comment, for offering an explanation, and acknowledging the prevalence of that particular cultural item.

You just attacked an explanation, including personal slag like "anyone who doesn't know".


I did it because you misquoted me about my use of the word "reverse" in a sentence, as I clearly pointed out above. So it was like making someone (me) look like they said something they didn't say.

But maybe I went at it too strong.


>Entschulding

Typo, entschuldigung.


My entire career - New Zealand, and Australia - asking questions is weaponised (as I stated above)

Graduate, Junior, Senior, Team Lead, - my title hasn't mattered to the response


You're working in toxic workplaces. Most of them aren't like this.

(I believe you when you say that most of yours are like this.)


I bet most workplaces are toxic in exactly this way.


Unionize


some of the descriptions above of thoughtful supportive work places where people are free to explore different ideas and question assumptions sound like paradise.


Or judged via rose coloured spectacles.


I think a lot about how much we altered our environment to suit cars. They're not a perfect solution to transport, but they've been so useful we've built tons more road to accommodate them.

So, while I don't think AGI will happen any time soon, I wonder what 'roads' we'll build to squeeze the most out of our current AI. Probably tons of power generation.


This is a really interesting observation! Cars don't have to dominate our city design, and yet they do in many places. In the USA, you basically only have NYC and a few less convenient cities to avoid a city designed for cars. Society has largely been reshaped with the assumption that cars will be used whether or not you'd like to use one.

What would that look like for navigating life without AI? Living in a community similar to the Amish or Hasidic Jews that don't integrate technology in their lives as much as the average person does? That's a much more extreme lifestyle change than moving to NYC to get away from cars.


"Tons of power generation?" Perhaps we will go in that direction (as OpenAI projects), but it assumes the juice will be worth the squeeze, i.e., that scaling laws requiring much more power for LLM training and/or inference will deliver a qualitatively better product before they run out. The failure of GPT 4.5, while not a definitive end to scaling, was a pretty discouraging sign.


Customer service will be almost fully automated, and human customers will be forced to adapt to the bots.


Or customers will find other providers who don't annoy them.


History so far suggests this is a dim possibility.


Exactly why Comcast and Google went out of business with their abysmal customer support.


It already has with IVRs . I wonder if as a generalization, current technology will keep being used to provide layers and layers of "automation" for communication between people.

SDR Agents will communicate with "Procurement" Agents. Customer Support Agents will communicate with Product Agents. Coffee Barista Agents will talk with Personal Assistant Agents.

People will communicate less and less among each other. What will people talk about? Who will we talk to?


We didn't just build roads, we utterly changed land-use patterns to suit them.

Cities, towns, and villages (and there were far more of the latter then) weren't walkable out of choice, but necessity. At most, by the late 19th century, urban geography was walkable-from-the-streetcar, and suburbs walkable-from-railway-station. And that only in the comparatively few metros and metro regions which had well-developed streetcar and commuter-rail lines.

With automobiles, housing spread out, became single-family, nuclear-family, often single-storey, and frequently on large lots. That's not viable when your only options to get someplace are by foot, or perhaps bicycle. Shopping moved from dense downtowns and city-centres (or perhaps shopping districts in larger cities) to strips and boulevards. Supermarkets and hypermarkets replaced corner grocery stores (which you could walk to and from with your groceries in hand, or perhaps in a cart). Eventually shopping malls were created (virtually always well away from any transit service, whether bus or rail), commercial islands in shopping-lot lakes. Big-box stores dittos.

It's not just roads and car parks, it's the entire urban landscape.

AI, should this current fad continue and succeed, will likely have similarly profound secondary effects.


Google Cal will auto-decline meetings that occur during ‘out of office’ events.

I have several of these ‘out of office’ events recurring in my calendar that I use for concentrated work


Up or out.

Yes, re the gamesmanship on pay, but if you don’t have the specific ability to bring in new business, then you’re on your way out, no matter how good a lawyer (or whatever) you are


Never underestimate a rat


The theatre is also there for the immigrants (or whatever undesirable that's being targeted).

It has a chilling effect on what people say and do.

You'll notice the same effect in other states that have armed people who turn up unexpectedly to make people disappear.


Literally manufacturing trust eh?


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

Search: