You realize that you just answered your own question? It’s bad enough when a manager also tries to be a developer working in a production system (instead of doing research, POCs, enablement type work if they really want to code), the last thing I want is some vibe coded slop from my manager.
But yes as a staff cloud architect who specializes in app dev, I very much treat AI as a junior developer who “learns” by my telling it to summarize discussions/preferences in markdown in the repo.
I do a phase approach when I use AI just like when I don’t where I test a little at the time. It gets too difficult to manage and explain otherwise.
I will tell you that “being a good engineer” doesn’t just involve “I codez real gud”, that’s the mark of a mid level engineer at any company that I’m aware of that has real leveling guidelines including my first hand experience at one BigTech company and my second hand experience at another.
Managing organizational complexity and people skills is required.
On a technical level I have always tried to carve out an “epic” or a “workstream” with clear interface boundaries between other work before I became a more senior “architect”
This is Dropbox’s leveling guidelines for instance
The press has never been believable. How many innocent people were beat, framed and shot and the press just took the word of the police? Rappers in the 80s were talking about police brutality. But no one believed them until the Rodney King video in 1992. Now many don’t instinctively trust the police because everyone has a camera in their pocket and publish video on social media.
On the other side of the coin, the press and both parties ignored what was going on in rural America until the rise of Trump
A government related alignment may lead to increased truth?? Have you been paying attention in the last year where the government is cleansing government websites of any facts that don’t support its narrative
Yes, I believe the reason we have got to this point is the destruction of institutions such as the press.
Historically the press had pushed narratives controlled by state elites who also had a vested interest in the state wellbeing.
Today these are pushed by foreign entities or the more extreme the more engagement.
That's why conspiracy theories replaced established truths, the populist left believes in anti-state slogans such as "defund the police" and the populist right wants to destroy the supreme court
AI alignment might return the elites controlled narratives which were apparently crucial for democracy
You realize you are being part of the problem that fell for the same talking points you accuse others of right?
The “leftists” were arguing to demilitarize the police and spend money on mental health programs and when someone is having a mental health crisis, send someone trained to help the person instead of trigger happy untrained police who don’t know how to de escalate
It is the same thing, it's taking institutions of the democratic state and dismantling them
exactly like trump's attack on the supreme court which could also be explained with excuses such as "a non elected institution is trying to curb the will of the people", and that's just the top off my head
No one said “not to have a police force”. They said not to have a militarized unaccountable police force. Three of the nine members of the Supreme Court were appointed by Trump and three other justices are conservative. They almost always rule in his favor and have given him unprecedented power
the slogan is "Defund the police" not "eliminate the police force militarization" or "realign police funding to mental health". It is an anti-institution rally call.
The fact that Trump has appointed Justices does not make Trump's attacks on the supreme court whenever they cancel any of his radical programs less real.
You are trying to argue that he shouldn't attack the supreme court, while I am arguing destroying institutions is ingrained into populism
I think they really should charge with micropayments, and they could even roll out their own currency for that if need be. Ads suck.
Actually, all the AI companies together should choose a micropayment system to focus on. I know in fashion, I've seen what would seem like competing brands center around a common "pillar of influence."
Also, if (long-tail?) AI companies work together, they could install appliances and terminals around cities. The most immediate use case - transport timetables. It seems like a no-brainer the more I think about it. Especially good for tourists who don't speak the local language. Governments may end up wanting to do that anyway and could subsidize the cost. It really depends on how fixated people are to owning their own screen, versus using someone else's. Those city screens could end up billboards anyway - especially for local businesses. They could print for a fee too and third parties could pay to get their app listed. Also, it's worth considering the increase in wealth inequality and rising hardware costs for people to own and stream into their own device. So this could be like the Internet Cafe 2.0.
Incidentally, there's a recent thread about someone streaming HN to a cheap display: https://news.ycombinator.com/item?id=46699782 - why not have such displays around town? I guess one major problem is vandalism.
I prefer dumb phones, and then prefer to not have to carry one 24/7.
Device lock-in is a whole other discussion. Why can't phones be switched off anyhow in terms of telco signal? Yet their Wifi and Bluetooth can be. Weird. What are they doing in stealth?
Well, I was thinking of integrating payments into the scrolling chat, the more you scroll down, the more payment increments you hit. It happens in the background, and there can be a tally somewhere on the screen of how much you are spending. Also, compute usage can be used as a multiplier.
From what I've read on HN, running advertising technology is an expensive and complex undertaking. I'd be trying to skip it altogether and keep the subsequent costs, intrusions and headaches away from users.
The other good thing about micropayments: being able to instantly divert some of the revenue back to content and training sources. That'd make it more righteous too and make it more conducive to cooperation from them (eg realtime pings.) Content will improve as a result, better justifying the costs. Could lead to less bot rampage too lowering bandwidth costs overall.
It'd also remove the temptation (hopefully) for AI companies to resort to black-hatting: scamming, backdoors and trojans to recoup their costs. That's seriously important and code-checking can become less of a priority and thereby result in time-saving for end users.
You don’t think your preference for not using a smart phone makes a viable market do you - seeing global penetration of smartphones is 90% and even higher among those who can afford to travel?
I think compulsory 2FA and the trend towards must-have downloadable phone apps is a problem, but that may not be fully evident... yet. In my experience, being tied to a phone and phone number is a problem. Also, when you carry your phone, your funds are at risk too from tech-jacking (as opposed to car-jacking) .. especially with crypto, right?
It's a personal choice - you are also tied to a battery charger. Wait, solar panels are getting better.
Why is it so difficult to run a mobile app on a PC? Why can't there be a device that I connect to my laptop to turn it into a phone (voice + texts) whenever the need arises? Weird. What's with the identification required at SIM point-of-sale? Is someone trying to track me or something?
I'd be more inclined to carry around a smartphone if everything about it was open-source and based on open standards. It's just another networked computer, really.
Oh. I do pay since basically the day they offered it. It's not a matter that they should care.
No ads is a point of product differentiation. One among many. But in some sense ads are a natural resource curse that pervade the whole company. Again, I point to Apple vs Google/Meta.
People who haven't seen an ad or paid a subscription in 20 years are still trying to figure out why no one listens to their opinion on how to make the internet better.
> Everything the user account posts immediately gets marked as dead
Yes, that's one of the standard ways that HN deals with troublesome users.
Seeing that sort of thing is the price you pay to see everything, when you turn showdead on.
In general, this sort of thing is probably better dealt with by email. My experience has been that the mods are very polite and well-reasoned in that channel.
With C++ formatting is optional. A better test case for LLMs is Python where indention specifies code blocks. Even ChatGPT 3.5 got the formatting for Python and YAML correct - now the actual code back then was often hilariously wrong.
I can't even get Github Copilot's plugin to avoid randomly trashing files with a Zero No width break space at the beginning, let alone follow formatting rules consistently...
I am the last person to say anything good about CoPilot. I used CoPilot for a minute, mostly used raw ChatGPT until last month and now use Codex with my personal subscription to ChatGPT and my personal but company reimbursed subscription to Claude.
A quick search finds many COBOL checkers. I’d be very surprised if a modern model was not able to fix its own mistakes if connected to a checker tool. Yes, it may not be able to one shot it perfectly, but if it can quickly call a tool once and it “works”, does it really matter much in the end? (Maybe it matters from a cost perspective, but I’m just referring to it solving the problem you asked it to solve.)
Clearly it isn’t just “broken” for everyone, “Claude Code modernizes a legacy COBOL codebase”, from Anthropic:
Taking Anthropic reporting on Anthropic, at face value, is not something you should really do.
In this case, a five stage pipeline, built on demo environments and code that were already in the training data, was successful. I see more red flags there, than green.
You mean the same Apple who will remove an app like Tumblr for a little consensual nudity posted by people and is too afraid of what Trump might say to remove X which is allowing none consensual undressing of women just by posting a picture and telling Grok to remove clothes - including CSAM?
But yes as a staff cloud architect who specializes in app dev, I very much treat AI as a junior developer who “learns” by my telling it to summarize discussions/preferences in markdown in the repo.
I do a phase approach when I use AI just like when I don’t where I test a little at the time. It gets too difficult to manage and explain otherwise.
reply