I don't disagree with anything said here, but it sure would be nice if Meta (and any other companies building social media applications) focused more on what's right for children, both short and long term, than the math of what is profitable.
The problem is that "what's right for children" should be better parental controls, and better enforcement of treating self-declared ages under 18 separately. Instead, the New Mexico AG and other assholes believe it means we need more mass surveillance, mandatory age verification, bans on encryption, and other insanity.
It's a bit of a tangent, but Meta actually really annoys me
Is the general war on encryption bad? Yeah
Meta? Meta forcibly upgraded people on Messenger. Now chats are broken. First they took away your friends https://news.ycombinator.com/item?id=4151433 and then they REALLY took away your friends, by doing this
If nothing else, the timing is suspect given the attention and press open weight models are getting over the last few weeks. The releases of Kimi and other models is getting open weight models enough attention that the US government, perhaps pushed by OpenAI and Anthropic, to think about taking action against these models for security concerns.
Good to see that more neutral companies (Microsoft and Meta to name two) are pushing back against US government involvement:
Yeah came here to say the same thing on the lack of absolute units.
Speaking as an older vinyl collector myself, with kids who have also gotten into it...it would surprise me if the total unit numbers were even within 2 orders of magnitude.
Vinyl just feels different. The feeling of holding the cardboard cover and the weight of the album on a turntable can't be replicated at all with CDs.
Agreed and it's unlikely to slow down that descent into slop coding until there are some visible issues with it. The market (i.e. companies) are going all in on AI coding because everyone else is doing it and the concern (understandably) is that if a company doesn't join that race they will lose because they never even entered the race.
The trick will be for companies to go fast enough to be in the race, not winning it, just in it. That will allow the time/space to let someone else, whoever is going fastest, to trip and fall so the rest of the pack can learn.
The tip and fall moment could come as a major incident (reliability and/or security) or loss of revenue because of bad products that customers don't like enough to use.
> There's a world of difference between managing a dozen or so individual contributors vs managing senior managers / directors.
I agree, but in my opinion, your point about needing to know how things work still holds. It's less relevant as you move up to managing managers and larger orgs, but it never goes to zero if you're going to be successful.
I think it holds true throughout if you scope it to knowledge about your reports’ domain. If you’re managing managers you should have a good handle on what a manager does and different ways to do it.
Will it be is a different thing though. And if it’s not, who exactly is accountable?
With funds and portfolio managers that run them, there’s a clear accountability model (if the fund sucks, the manager loses their job and the company loses credibility)
With AI agents doing the management, who is accountable when the fund sucks? If it’s the customer, we’ve moved accountability from someone who at least in theory, knows what they’re doing to someone who has little to no clue.
You have to be accountable for what you have the model do on your behalf. I hear what you're saying, but there are also issues with the hedge fund accountability model. There are certainly swaths of fund managers who are only there because they got lucky or had the right pedigree, and more that are better traders but never became a fund manager because they got unlucky or had other passions.
An individual investor can invest with their risk appetite on their time horizon and not be subject to Citadel's "5% draw down in a quarter and you're fired" culture which can be toxic to returns over time.
Per a report that came out the other day, the GitHub move to Azure has been slowed down (i.e. I don't think it's done). But maybe you have newer/better info than me
Yeah, that and Microsoft has been slow to move the infrastructure to something that scales better to handle that load.
The more surpassing part is that Microsoft hasn't figured out a way to manage/contain the AI-sourced traffic better so it doesn't create all this noisy neighbor problems for non-AI usage/users.
Github's core platform doesn't really make that separation, anything a human can leverage on github an AI agent can as well, just faster and with heavier usage. End of day agents and humans are using the same services.
Sure, still need to enable access the same info but feels like bucketing the clients into
bucket1 = clients that were working just fine before (users and whatever automation they had in place)
bucket2 = ai clients that contributed to, if not flat out caused, the scale problems
then slowing down/limiting the bucket2 clients while keeping the bucket1 clients rolling as-is, is both doable and keeps existing customers happy while the underlying infra gets scale/perf improvements needed to support ai clients at scale.
> Every use of AI for these robs the employee culture of a genuine trust building moment.
Spot on.
The erosion of communication and relationships between people in the workplace (or even outside it) that AI contributes to is something that we don't talk about nearly enough. Society today has already suffered greatly in these areas thanks to social media, and AI just makes it worse.
People (in general) are really struggling to understand when/how to use AI to be more productive and happier (and imo there is a way to do it, by offloading the grunt work to AI). With the constant rush and jamming of AI down everyone's throats though, its hard to be able to take that step back and think "is this use of AI making me happier/more productive".
Yes. The role of good management here cannot be understated. Good management (all the way up) is the difference between saying "be more productive, here's an AI subscription" and people understanding what types of usage are actually wanted and useful.
As it is now, with just the vague handwaving many managers are doing, people are hearing "You should reach for AI immediately anytime you get an input that you technically can paste into the AI" - so we can't be mad at them if they're just doing what they think they're being told to do.
I was a bit annoyed with an IC recently because he would just respond to peoples messages with "Heres with Claude says".
He didn't understand why I was so annoyed because we have been pushing them to use AI for everything.
I had to explain to him, its like when google first came about, yes you can just send people Let me Google that for you links, or you can respond politely and reasonably to build a relationship. In the end I don't think he got it and it's something I'll need to continue working on with him.
Where are you guys working where people are doing this? I work in a company where leadership is also ramming AI down everyone's throats, but I don't recall ever getting copy/pastes from LLM as responses to E-mails or chats. My biggest problem is people not reading/answering their E-mails and chats at all, or finally getting back to me long after the due date of whatever I'm asking about. Which is a different workplace comms problem altogether.
Design docs on the other hand have been fully taken over by the slop machine. They all kind of look the same now, and give off that familiar "I didn't write it so you might as well not read it" vibe.
I was working in an env where I started to suspect that people I was chatting with were using LLMs. These were people that didn't want to talk to me either way, so there was not much lost here. I suspected that, because the technical expertise they were showcasing when responding to messages would evaporate when talkin f2f
Norms surrounding the use of LLMs are in the process of being established, it's a new frontier. Many people rely on these signals over common sense. The feedback loop will lead to corrections in time, for now people are sussing out where the boundaries of appropriate-use are. Corp/gov policy is still lagging as well.
I really am not a big fan of this... Hand-waving, I guess? Around this problem. Saying "well the norms are still being established" feels kind of like a "well don't really get mad at the people doing it, they're still trying to figure out the boundaries of acceptable use" kind of thing to me. People should already know that this kind of behavior is unacceptable. The fact that they don't is very, very telling and says a lot about the people doing it IMO.
They should, but they (some) don't. As with most things, it's more attributable to ignorance than malice. Not much I want to do with what this "tells" me.
Maybe. It doesn't help that a lot of corporations are pushing their employees into dark patterns around LLMs. That in turn informs their own personal use of LLMs outside the workplace
