Hacker News new | past | comments | ask | show | jobs | submit | foobiekr's comments login

Why is this downvoted? It’s clearly a reference to the zizians and MIRI and Bayesian nutjobs is an absolutely correct assessment.

Stuart Russell and Geoffrey Hinton have both expressed concerns that AI could lead to human extinction and neither are nutjobs

I wasn't explicitly referring to the more "sane" people expressing doubts regarding Ai.

Hinton at least says that other issues in Ai should be dealt with, rather than just being an Ai doomer who only fears Ai takeover he actually realizes that there are other current issues as well

At this point, how many times should we have been dead for eliezer?


Like almost all the other doomers, Eliezer never claimed to know which generation of AIs would undergo a sudden increase in capability resulting in our extinction or some other doom, not with any specificity beyond saying that it would probably happen some time in the next few decades unless the AI project is stopped.

Idk, a few years ago when chatgpt came out he was saying things like "if we're still alive in 3 years (...)" where chatgpt 3.5 was still a glorified transformer. And modern llms still are. It's the constant fear mongering that stings my nerves.

And well, I'm not surprised nobody knows which generation of Ai could undergo an increase causing our extinction, it's not even sure if there could exist such a thing, let alone know which generation


He has been saying for a couple of years that it is possible any day now, but in the same breath he has always added that it is more likely 10 or 20 or 30 years from now than it is today.

It is not a contradiction be confident of the outcome while remaining very uncertain of the timing.

If an AI is created and deployed that is clearly much "better at reality" than people are (and human organizations are, e.g., the FBI), that can discover new scientific laws and invent new technologies and be very persuasive, and we survive, then he will have been proved wrong.


Ok, I think I might get a heart attack sooner or later, it's a possibility, although not very high.

If I said so, you might ask me if I saw a doctor or something similar to make me suspect that, and that's my issue with him. He's a Sci fi writer that's scared of technology without a grasp of how it works, and that's OK. He can talk about what he fears, and that's OK. It still doesn't mean we should take him seriously just because.

My pet peeve is that when trying to make laws regarding Ai - at least in Europe - some considerations were done regarding how it worked, what it was (...), how it's being talked in academic literature. I had a lawyer in a course explaining that, and while not perfect you eventually settle on something that more or less is reasonable. With yudkowsky, you have a guy that is scared of nanotech and yada yada. Sure, he might be right. But if I had to act based on something, it would look much more the eu process to make laws and less the "Ai will totally kill us from now to 30 years trust me". Perhaps now I'm more clear

And don't get me started with the rationalist stuff that just assumes pain is linear and yada yada


Eliezer has written extensively on why he thinks AI research is going to kill us all. He has also done 3-hour-long interviews on the subject which are published on Youtube.

And perhaps YouTube is the appropriate place to talk about the probabilities he pulls out of thin air such as chatgpt 5 killing us with less than 50% chance, the badmath he showed to us a few years ago on reddit, and him proposing to trust bayes rather than the scientific method

At least he could learn to use some confidence intervals to make everything appear more serious /s

I'm very much in favor of research in Ai safety, maybe done with less scare and less threats of striking countries outside of the gpu limit agreement (and less bayes, God)


That's more or less what I was referring to, but I should have been more clear

Keep on telling yourself that.

AM for example could plausibly be trained on 4chan and war footage.

Most of the criticism on display here is the outrageous, implausible lies that the tech industry leaders are telling to stupid people who believe it for propaganda purposes to avoid regulation and scrutiny.

None of the bullshit coming out of Musk, for example, is real, it’s not even plausible, it’s just lies for dumb people.


No you

They really aren’t. And I know a lot of them personally.

Ah, McKinsey. That their pro-remote study was bogus will probably come as a shock to the four people left on planet earth who aren't familiar with "where there's fraud there's McKinsey"

Why do you think "DEI people" are running DEI programs?

Scott Alexander has a good summary of what went wrong with the former mindset (fill the job with the qualified candidate becomes mimic the population at large or else): https://www.astralcodexten.com/p/book-review-the-origins-of-...

The problem with DEI is that it did, in fact, turn into a policy of racial quotas, only the quota-ness was denied even though the threat of legal action was omnipresent.


I work for a company that made a big deal about how well it worked. The CEO and chief people officer both made a lot of public statements about it, during and just after covid. They did this claiming they had the numbers to back it up and said numbers but they basically made them up. It was PR and they were using it to aid recruiting into what was and is a company that young people avoided. What numbers they did have were mostly inertia and a side effect of lockdowns (people had little else to do); new projects and new teams floundered very visibly.

Internally, it was very clear that it did not work. The numbers were massaged to back up the executive leaders, but everyone was pretty clear what was going on, and even the exec leaders, in leader-only meetings, did eventually admit that it was "more nuanced" initially to "does not work" internally. The company has since moved away sharply from remote employees. It's still not full-on RTO, but it's edging toward it.

I'm not saying this is everyone, but I think people should really take the 2020-2023 rise in remote and the narrative around it with a grain of salt. Most of the companies that championed it have reverted, and they aren't doing that because bosses are control freaks.


This often just feels like bad management.

They go remote, but don't change a lot of other things or attempt to mitigate the downsides (there are downsides, everything is a trade off) and then claim its a failure when they need a stealth layoff.

Also IBM has a long history of "Resource Actions" so this type of thing is not all unexpected from them.


I don't know if it's bad management per se. I think some people are very well suited for remote; some people aren't. Probably a rough extension of introversion/extroversion in the people mix.

If you take a bunch of very extroverted people and have them all work remotely they will not have a good time (in general).

Equally; if you take a bunch of very introverted people and have them in an office they'll really not like it, especially in open plan.

The other problem is fraud levels in hiring for fully remote is absolutely shocking. There are so many stories now of fake candidates etc, massive cheating in interviews with AI, etc. I've seen many stories like that even with really 'in depth' interview processes, so much so people are now going back to in person interviews en masse.

My rough take is that organisations need to really rethink this home/office thing from first principles. I suspect most engineering teams can work as well/better fully remote. I very much doubt all roles are like that. I think we'll see WFH being based on department or role rather than these global policies.


I think both problems are real, and your last paragraph really gets at why. People and jobs vary widely, and so does the quality of management. If you have a strong business and reasonably mature teams, you might not even realize problems with your management culture and practice are until something big changes; conversely, if you have strong managers you might have been able to soak up a lot of personnel and job issues before they got attention because some unappreciated middle managers put a lid on potential problems first.

In all cases, you really someone with time to look at the business as a whole to evaluate these things. For example, one of the things which has made RTO unproductive for many workers are open plan offices, which is a really easy problem to see and fix if workplace productivity is someone’s job but not if the RTO push is being driven by politics or the need to justify leases.


Agreed, but falling commercial real estate prices will allow some forward thinking companies to go back to private offices for developers, which arguably is the best solution (apart from commute/flexibility) for most, productivity wise.

Funny you mention fraud... I worked for a company for quite a while that was absolutely dedicated to WFH for engineering - but swore up and down that sales just couldn't work without "bullpen" office setups.

Come to find out at least one entire office was engaged in widespread misreporting and fabrication. Turns out fraud is pretty tempting when you can easily avoid any paper trail.


yeah I feel like blaming management is like blaming teachers when students got bad scores during remote-schooling. You can give them all the resources they need to succeed, but if they'd rather go to the dog park in the middle of the day, there's not much that can be done.

Meanwhile, it's worth noting that some students excelled at remote schooling. But most are reading at a level 3 grades behind.


The metaphor breaks down a bit when you consider that teachers don't generally get to pick their students while organizations get to choose their employees. Failing to choose the right employees is a failure of management.

Unfortunately many of the employees _most interested_ in remote work are such because they want to do things other than work.

Not all. I work with some remotes who are awesome. But the 24 year olds who want to work remotely from Thailand aren't getting their shit done.


Re:introvert/extravert I suspect it's the reverse.

Extraverts, broadly, aren't afraid of picking up the phone and calling you to chat about the email they sent you three minutes ago while driving and also on mute in a zoom; introverts can use remote work to be unreachable in a way they can't if you can just walk over and impose yourself on them.


Seeing posts on here and Blind advocating for interviews to go back on-site due to cheating next to "RTO bad" posts is wild af.

It's wilder still that the handful of times I've dealt with this have all been before RTO!


I don't mean completely in person, but I do expect a lot of companies will want to meet the person in real life at least once. Which adds huge logistics problems.

Btw I'm not saying 'cheating', that's one thing. I am meaning industrial scale fraud with remote candidates. Eg having one person interview then another (much worse) person gets the job. There are gangs that are going to almost unbelievable lengths to do this.


About a year ago I moved to a new, largish company and, for the first time in my career, got to see how a company can be bad at remote work.

It's by being bad at work, period, but in ways that can be partially mitigated by being in-person. Poor documentation of processes, lots of know-the-right-person involved in getting anything done or figured out, using Teams (its design is remarkably awful for organizing and communicating within and among... teams) rather than literally any other notable chat system, et c.

This stuff is also making in-person work less efficient but it's easier to work around the problems when in-person.

Better than resisting remote work, would be for them to suck less at managing a business. Even if they continued resisting remote work, they should do that!


And being in an office doesn’t help if no one on your team is in the same office. If you work in a large company that has multiple offices, you are still going to have the sane problem because eventually the person you need is not going to be in your office.

Even the small companies I’ve worked at (100-700 people) had multiple offices where you had to coordinate time to meet with the people you needed.

I’ve also worked remotely for the second largest employer in the US. Amazon has internal “interest” channels for each service team (the team responsible for an AWS service). Anyone could ask a question and usually one of the developers of the service would help.


> rather than literally any other notable chat system, et c.

What is better? I hate Teams, but Slack really wasn't much better.


Teams' core problem is that the actual Teams-section is more like a bulletin board than a chat system, almost like it was targeting that weird impulse companies had for a few years to build "company facebooks" or whatever.

The real chat part is cordoned off in ad-hoc channels that individual users can sticky, but that aren't "structural" and can't really have order imposed on them, if that makes sense.

It's like if Slack only had the DM and group-message feature, and no channels.


No arguments about the crappiness of Teams.

My thing is that while better IM systems exist, none is what I would call "Good" or even "Acceptable". Being better than Teams is not really saying much :-)

I haven't used Slack in years, so I can't speak to it, but it sucked when I used it. Back when our team was all colocated in one building, I intentionally had my IM app turned off and disconnected. Interruptions in person suck, but with Slack et al interruptions were multiplied significantly. Kind of: "If you can't be bothered to get up and walk to my cube, it probably wasn't that important."

What I want from Teams and similar SW:

A way to, with a keystroke, mark all messages as "Read" (even when focus is not on the window).

A way to, with a keystroke, print out all unread messages on my console (or in a popup window, or whatever).

In other words, just give me a damn API I can program these things with. Teams' API lets me get messages, but will not let me see if a message is read or unread.

Any app that forces me to open up the window, click on a dozen channels to read all the latest messages, sucks. Period. I should be able to read it all with one click/keystroke, and have them marked as "Read" when I do it.


Wait, what? We're moving to Teams soon...

This looks like channels do exist, is it new (there's no date on the page) or do they not work as you'd expect? https://support.microsoft.com/en-us/office/the-new-chat-and-...


At Google, they found that engineers L5 and above got more work done with RTO, and engineers at L4 and below got significantly less work done. WFH is great but it doesn't work for fresh engineers (who are often the most gung-ho about it as well).

Did you mean L5 and above got more work done with WFH? Since the next sentence implies that it was the fresh engineers who were most impacted by WFH.

I think it goes beyond bad management.

These are my disadvantages of working remotely. I say all of these things as an advocate for hybrid work arrangements and co-working spaces/satellite offices:

1) Some people work better in an office. Offices are literally designed for working anyhow.

2) Some people didn't, and/or still don't, have optimal conditions in their house to work remotely.

I've seen tons of people on camera (another thing some management likes to "encourage" by mandate) who are working out of bedrooms, closets, or other makeshift rooms in their house. This is just _asking_ for a constant barrage of distractions.

3) Some jobs aren't compatible with remote work. Examples:

- Tech sales (moreso for complex sales and expansions than new sales)

- Many people who work in the public sector (even before this administration's aggressive RTO campaign)

- Most folks doing hardware or embedded work

- Pretty much everyone that we interact with outside of our home on a daily basis, like front desk personnel, doctors, mechanics, retail and restaurant staff, etc.

This creates an unfair imbalance of "haves" and "have nots". It is also very easy for the "have nots" to typecast those who WFH as lazy, especially given some of the memes of people doing all sorts of other things during core hours.

4) Some people don't naturally communicate what they're doing over Slack. This is the one thing I'll blame on management is communication.

Weekly "15-minute" hour long standups and check-in meetings covered for people like this back when we worked in offices, but it can be easy for these checkpoints to slip in when everyone's remote.

Now, these meetings existing are, in and of themselves, signs that management can be improved. Between Slack/Teams/whatever, bug trackers, Git commit histories, Office 365/Google Workspace APIs and all of the other signs of life of people doing things, there are ways for the PHBs to check that people are doing things so that they can report the things being done to their PHBs so they can report to their PHBs all the way up to the board and investors.

It would be great if more companies invested more in their processes to make it possible to assess productivity without needing inefficient meetings. This would make it possible to be a high-performing company regardless of location.

But change is hard, and it's easier for senior leaders/execs to throw their hands up and say "this isn't working; back to the office, now", especially when those leaders are already traveling all of the time as it is.

(I know that the trope of CxOs who golf/eat steak dinners all of the time is common; my experience working with people at these levels does not completely reflect that.)

5) Work-life balance is so much easier to immolate when working remotely.

When your home is your office and your work apps are on your personal phone, it takes the mental fortitude of a thousand monks to not be "terminally online" at work.

"I'll just hop back on after I'm done with the kids/dinner/etc." is the new normal. It existed before WFH, but it feels so much worse now, as the technology needed to set this up is so much more pervasive (mostly MDM being mature for Apple devices and Android becoming much more secure at the cost of everything that made Android fun for us hackers).

This has the fun side-effect of making people who try very hard to keep work and life as separate as possible look like slackers even when they're not.

6) Establishing rapport and camaraderie is much harder to do remotely. This "just happens" when you're working next to the same people every day for months/years at a time.

This was most evident when I joined a new company after COVID to avoid an acquisition. Almost everyone was super tight with each other because they hung out all of the time. There were so many inside jokes/conversations/memories that I was basically left out of, and because traveling was impossible then, forming new ones didn't really happen.

I get that many on this board view this as a feature, not a bug, but friends at work is important to some (most?) people. It's the one thing I miss from the before times more than anything else. Well, that and traveling all of the time!

7) Last thing I'll say on this: onboarding, in my opinion, is much worse when done remotely.

I've switched companies four times since COVID. ALL of these onboarding experiences have had some combination of:

- Loads of training materials, like labs and new hire sessions, that are dry as toast over Zoom but can be extremely engaging in-person,

- Some kind of buddy system that falls apart because everyone is drowning in a sea of Zoom meetings and the last thing people want to do is have ANOTHER zoom meeting explaining things about your new job that are kind-of difficult to explain without shadowing, and

- An assumption that you are a self-starter who will learn how to do your job by self-organizing meetings with people and scouring whatever documentation/knowledge/recordings/etc you can find.

This might just be a 'me' thing, but I've found remote onboarding to be a poor substitute for onboarding at an office somewhere.


> I've seen tons of people on camera (another thing some management likes to "encourage" by mandate) who are working out of bedrooms, closets, or other makeshift rooms in their house. This is just _asking_ for a constant barrage of distractions.

At home, there have never been more than three other people in my house, when I’m “at work” with my door closed, they knew not to bother me. At work in an office there are constant distractions.

As far as “tech sales”. I’ve lead my share of complex cloud tech projects from discovery, customer acceptance to leading the delivery - all remotely. Yes sometimes I had to travel to the client’s site. But I haven’t needed to be in the office with the people on my team (who were sometimes in another country).

My coworkers are just that my coworkers. At work, “I’m taking a step back to look at things from the thousand foot few”, “taking things to the parking lot”, and “adding on to what Becky said”. I’m a completely different person at home. At the end of the day, my “friends” at work are not interested in keeping their jobs. I go to work to make money - not friends.

I’ve worked for two companies remotely since 2020 - Amazon and now a much smaller company. They both had excellent onboarding procedures. While AWS wasn’t “remote first”, my department (Professional Services) was as is my current company. Both had “onboarding buddies” and Amazon had a list of people you should set up 1x1’s with an instructions for the relevant internal systems you should use.


I think we are going to agree to disagree on some things, but I understand that I am a professional weirdo when it comes to the WFH/RTO battle. I'm a child-free late-30s guy who has always loved commuting (traffic and all) and working away from home and treats the airport, airplane cabin and hotel room in some other city as a collective happy place.

> As far as “tech sales”. I’ve lead my share of complex cloud tech projects from discovery, customer acceptance to leading the delivery - all remotely. Yes sometimes I had to travel to the client’s site. But I haven’t needed to be in the office with the people on my team (who were sometimes in another country).

I was a cloud/DevOps consultant/SA as well before I moved into tech presales. It's a different world, even though it doesn't seem like it would be on paper.

Delivery can be (and usually is, these days) done remotely, but I've found that finding new opportunities to expand or sell into new parts of business is easier when done face to face. The human part of the job is difficult to replicate over Zoom, in my experience.

That said, when I was a consultant/SA, I much preferred pairing with clients in person than over Zoom. I enjoyed the travel and found sharing a keyboard to be more engaging than talking at a screen for hours on end. I realize that this was probably a "me" thing and that others are totally fine with remote pairing.

> My coworkers are just that my coworkers. At work, “I’m taking a step back to look at things from the thousand foot few”, “taking things to the parking lot”, and “adding on to what Becky said”. I’m a completely different person at home. At the end of the day, my “friends” at work are not interested in keeping their jobs. I go to work to make money - not friends.

This is where we differ. I'm at my best when I'm working with others in-person towards a common goal. While I'm also motivated by money and am not pining to make lifelong friends in the workplace, I miss going to the bar at the end of a long week and decompressing with others who "get it." My wife has this, and I'm always jealous about it. For me, doing this over Zoom pales in comparison.

However, all of this is why I prefer hybrid arrangements that are mostly remote with budget for monthly team get-togethers. I don't think being on-site every day is effective, but I've found being perma-remote to be really isolating.

> ’ve worked for two companies remotely since 2020 - Amazon and now a much smaller company. They both had excellent onboarding procedures. While AWS wasn’t “remote first”, my department (Professional Services) was as is my current company. Both had “onboarding buddies” and Amazon had a list of people you should set up 1x1’s with an instructions for the relevant internal systems you should use.

These are excellent systems _if you are a self-starter and know what you're doing_. They fall apart if you are junior that develops best in a dedicated environment, or if you prefer a more "social" way of onboarding.


I’m not denying the importance of sales meeting clients face to face or even leading the delivery side with a few face to face meetings during both discovery and turn over. I am one of the psychopaths that loves business travel and meeting clients in person.

I’m saying that it is silly to have those roles be in the corporate office of your employer.

AWS ProServe has a 3 month training program for their early career hires and career transitioners. It was all remote.

Even for their more senior roles, they had “AWSome Builder” where you had a two month “project” simulating a real world engagement where you had a mentor and five people from the department acting like stakeholders - CTOs, CFOs, directors etc. This was also remote where you had to do presentations.


Okay, AWSome Builder sounds really cool!

> Internally, it was very clear that it did not work

If it was clear, they shouldn't have trouble showing the data. Otherwise it's a case of "The data shows X, but my gut clearly shows Y"

I can certainly believe it didn't work for some companies/roles. But the burden is on the company to demonstrate it.


The company was publicly lying that it did because execs thought it would help us recruit younger people.

If they were willing to lie to juice their metrics during COVID, why would any outsider believe them now?

after covid is an illusion, there is no such thing as "after" for an ongoing pandemic....

and the company you work for seems rather incompetent


Unless you are willing to say there is an ongoing flu pandemic then I can't really agree with you.

Except one doesn't get the flu every 6 months like most people seem to get covid ; at most one gets the REAL flu every 10 years.

SOURCE: several studies and my own conclusion observing friends...


I caught covid once since it started.

> after covid is an illusion, there is no such thing as "after" for an ongoing pandemic....

The US death rate from Covid in Q4 2023 (so, pre-Trump II) is roughly the same as the US death rate from influenza. 14.9 or 16.8 deaths per 100,000 for Covid (first number is 12 months ending with Q4 2023, second number is the three month rolling window), 13.5 or 15.1 deaths per 100,000 for influenza.

"After covid" is a perfectly fine description of the current state of affairs.


I don’t understand how that conclusion follows from that comparison. We also wouldn’t describe the current day as “after the flu”. Both are endemic.

The phrase you picked to describe them is "endemic", rather than "ongoing pandemics". There's a reason for that.

by that token; you can argue HIV is endemic and nothing should be done about it... people will understand wearing condoms but apparently can't understand masking... hmm... neither are very difficult to do, I'd argue that wearing a condom is more tricky in some aspects.

the reason is simple: wearing a mask reminds people there is a risk and it doesn't "look good", despite its many advantages for many pathogens and pollutants; also acknowledging indoor air quality is crap is hard for businesses (and governments) to do since they would need to invest into better ventilation and filtration systems.


you are accounting for immediate deaths; you have to keep in mind that covid causes immune dysfunction, heart disease etc; 10% of people are still not recovered after 4 years...

long term effects are rare in the (proper, confirmed flu, not your cold you think is the flu) flu


> Most of the companies that championed it have reverted, and they aren't doing that because bosses are control freaks.

Funny joke. I needed a laugh.


And you believe this?

Surely then they have no swe reqs right?


just stating facts.

The fact being that Sarah "CEO of Nextdoor" Friar stated they have this A-SWE vapor thing? I don't see how it pertains to what the said being factual.

Yeah and respectfully the facts as stated leave the impression that she's full of shit and/or doesn't know what she's talking about. What she's saying is pure hype for investors probably. Vaporware.

Nextstep used compositing in the late 1980s.


[[citation needed]]

NeXTstep used Display Postscript. AFAIK this has no concept of 3D acceleration or anything akin to it, and was almost entirely unaccelerated.


compositing has nothing to do with 3d acceleration.

NeXT slabs and cubes did have a low-function blitter used to composite the main display.


Edit: this was the Turbo units, which did have a bus-assisted block copy.

Before that, just well-tuned 68030 burst block copy - they rendered to backing store and did block copies with a special case of vram to vram block copies (used for window compositing).


In this direct context, I absolutely think it does and it's integral to the definition.

What do you understand by "display compositing" that doesn't involve 3D hardware?


What the hell does 3D hardware have to do with the block copies compositing window managers like MacOS are doing?

Join us for AI Startup School this June 16-17 in San Francisco!

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

Search: