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

How about half of a flip phone?

Usually it starts with a locked down laptop, then request after request, you get more rights, even admin on the laptop because otherwise you can't do anything as a software developer.

Our (state) agency uses separate computer accounts for regular use and admin use. The admin account has neither network nor email access in order to prevent using the admin account on a regular basis. With state government agencies continually attacked with ransomware, we aren't as bondage and discipline as banks (which I have done before) or DOE (which was very severely extremely locked down and I was not connected to classified stuffs).

I have not needed admin credentials for regular development purposes since VB6 (which I was using until 2013). VB6 and ActiveX dlls required read/write access to the registry all the time.


Most banks give you a crippled version of Admin rights via a 3rd party tool. If you need more than that, tough/you're doing it wrong.

Banks, at least in the U.S., are highly regulated and this would be very abnormal and subject to regulatory non-compliance, especially when the bank crosses into “big bank” territory. This stuff is regularly audited internally and externally.

Even if you do need admin for a developer laptop, that has no bearing on the PAW you use to authorize deployments and otherwise interact with production systems. Assuming your employer is even a quarter competent (i.e., not Binance).

You can dev without admin.

No, it depends on the role.

Because they are treated very different. They are part of the unique political party that runs the company, like in China. Democracy exists only outside the company, inside the company it's the same political system as in China. What that means is that loyalty is the only thing that matters for those roles. Productivity not at all.

Maybe consciousness itself is also just a "pattern matching algorithm at scale". Maybe there is no magic.

How bout we think about consciousness as a VM; that VM can load one, many, several types of context processing. The environment (nurture) is what trains/expands the different ones and potentially unplug the other ones, etc.

One of those types of consciousness processing is indistinguishable from what an LLM does.

Everyone knows someone who seems to just call and respond based on whatever popular media of their choice of culture is on about.


I think it's an algorithm, just not a simple transformer pattern matcher that is gigantic and is brute-forcing intelligence.

I use AI so much that I too started hearing myself using the same phrasing as AI... not happy about it.

same here :)

I worked in the industry. It would have been really hard for China not to do better. Like really hard. They allocated 2 years, 150 people for a device that already existed but needed to be adapted to a different customer. It took 2.5 years. No improvements added. Every single manager in those companies is very political, no skills, technical or otherwise. So if chinese hired 20 competent people, gave them training on the industry's specific tools, they are already ahead.


>I worked in the industry. It would have been really hard for China not to do better.

Same. Major German car company I used to work for was so incompetent, slow and inefficient that it was inevitable they'd get shown off by Chinese cars, since Tesla did it before so it was a matter of time.


But even Teslas are 2-3x the price of equivalent Chinese EVs.

Apparently a Tesla model 3 has about 10-15 hours of human labor in it, so higher wages can't even begin to explain the price difference.


>But even Tesla's are 2-3x the price of equivalent Chinese EVs.

Well Tesla want to turn a nice profit.

Chinese EVs are sold on thin margins to capture marketshare.


There was a time when you would get a truly random password sent, every X months.


Sorry, I might be too young to remember those times. Can you tell me more, please? I'm genuinely curious.

Because from the sound of that, it feels that it would patch the `123456` problem, but opens up a new vulnerability - the password is known / being sent through / printed, so it can be leaked from that source.


Automated regular password updates were often stopped because it leads to people writing them down, usually on a sticky note on the bezel of their monitor. Or, if it's just an update requirement then you get PasswordOct26, etc.


It was just that. And email with the new password. Hard to remember at first but then, somehow, I could type it from memory. Not sure how they got rid of the emails after they sent them but there are ways. It can be leaked from the source, true, but people with better technical skills were handling that source. So I guess overall it was better. The big insight might be that you can't evaluate the security of a system only from the technical aspect. You have to take into consideration also the human aspect. And that's a very weak point. Even Communism failed because they didn't take into consideration the human aspect (people don't want to work more if they get what everybody else gets anyway).


Facebook will succumb to spam, like all previous social networks.


..so there's a kill switch. I see.


And if you do provide feedback, they also collect the session. So it's a way for them to collect prompts, answers and overall grade for how good the answers are.


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

Search: