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

Have you never run a team of software engineers as a lead? Agentic coding comes naturally to a lot of people because that's PRECISELY what you do when you're leading a team, herding multiple brains to point them in the same direction so when you combine all their work it becomes something that is greater than the sum of it's parts.

Lots of the complains about agents sound identical to things I've heard and even said myself about junior engineers.

That said, there's always going to need to be people who can reach below the abstraction and agentic coding loops deprive you of the ability to get those reps in.


People say this about juniors but I've never seen a junior make some of the bone headed mistakes AI loves to make. Either I'm very lucky or other people have really stupid juniors on their teams lol.

Regardless, personally, there's no comparison between an LLM and a junior; always rather work with a junior.


I've wrote this a few times, but LLM interactions often remind me of my days at Nokia - a lot of the interactions are exactly like what I remember with some of their cheap subcons there.

I even have exactly the same discussion after it messed up, like "My code is working, ignore that failing test, that was always broking, and I definitey didn't break it just now".


> Have you never run a team of software engineers as a lead?

I expect juniors to improve fast to get really good. AI is incapable of applying the teaching that I expcect juniors to internalize to any future code that it writes.


Nice straw man.

Nobody is talking about banning anything, we’re talking specifically about holding social media companies accountable for marketing to children a product that is knowingly addictive and potentially harmful to their health.

Part of the issue with social media is that no reasonable parent lets their 12 year old watch porn or drink but Instagram and ticktock are on a lot more 12 year old’s phone’s than you realize. Social media has network effects and creates tremendous social pressure to not make your kid “different” when half the classroom is sharing TikToks.

I’m not conservative in the slightest but I see no reason to treat social media any differently than alcohol, tobacco or gambling. Available without restriction to adults but limited to children under a certain age.


This stuff is still unclear to me. The addictive drugs, ones that punish a quitter chemically, are not mysterious, but gambling addiction certainly is. "Dopamine" won't work as an explanation - for instance I was once hooked on building a wooden table, which sucked up two months of my free time and lots of money, and damaged my thumbs, and no doubt I was driven by the dopamine rush of learning through the repetitive process of chiseling. But gambling is assumed to be a glitch, not a wholesome obsession. In what way does it differ? The addiction is very old, I'm sure there are accounts from the 1700s, and it doesn't even require a house to reel the gambler in - it could all be about informal games and wagers, still leading to huge debts. It's tempting to blame it on dumb ideas about luck and fate, but the dumb ideas involved could be varied and complex.

That's similar to dumb ideas involving social pressure. When people have a tendency to be dumb about a thing we use the law to restrict the thing, apparently. But this involves, in effect, an authoritative declaration of "that's dumb" by law. I feel personally threatened, then, in activities such as my woodwork, which might have been an equally dumb obsession! I know nobody's at all likely to regulate woodwork, but that's only because it's relatively unpopular. I could imagine a parallel universe where woodwork (portable somehow) becomes a trend that makes a young person feel socially relevant, and then it gets regulated. I think I disapprove of this interference with people's dumb notions.


This is no longer a matter of adults or minors; this is a matter of terrorist acts committed by a satanic cult and organized crime. And the response will not be limited to legal means; we will seek to respond with the same kind of terrorist tactics.


I think people are overreacting about what amounts to a patent. There's no evidence if BMW will actually use this and if they do how.

I'm not defending BMW here but there was a similar freak-out a few years ago about heated seats requiring monthly payments that ended up being a giant nothing burger.


There’s a million reasonable situations where this pattern could arise because of you want to encapsulate a domain behind a micro service.

Take the simplest case of a CRM system a service provides search/segmentation and CRUD on top of customer lists. I can think of a million ways other services could use that data.


I don’t think renting or buying makes a difference in communities. I had a much better relationship with my neighbors when I rented vs now where I own my own home.

I chalk up the difference to the fact that a lot of people my age couldn’t afford to live where I moved so my neighborhood is full of wealthier older people who I have nothing in common with.


I agree with you 100% as someone who's developed on and run many Kubernetes clusters over the years. Picking a choosing what to use can be daunting.

That said, the path to ease of use usually involves making sound decisions ahead of time for users and assuming 99% will stay on that paved path. This is how frameworks like Spring and Rails because ubiquitous.


I am a graduate of Khoury and find this very disappointing. I wasn’t a huge fan of learning Scheme/Racket but after looking back during my career I’m grateful for the education I got there.

This feels like a step backwards.


I have no idea why anyone would use Cognito unless they don’t care about availability.

Almost every other SaaS vendor supports multi-region active-active and Cognito does not.


> Almost every other SaaS vendor supports multi-region active-active and Cognito does not.

Who are we talking about here? Google and Azure?


auth0, okta, ping identity, azure, google


Had some pretty negative experiences with pricing/"enterprise" sales tactics by Okta (which now owns Auth0, and they used the same tactics on both products). I will take AWS pricing shenanigans over that any day.


I'll take the scummy sales tactics over the cognito API any day of the week


Given the choice between a crummy API and being driven bankrupt by a SaaS vendor, I prefer a crummy API. I suppose your calculus might look different if you have a lot of money or an employer with great negotiating leverage.


Okta has been plagued by security issues [1], never heard of Ping Identity, Azure only makes sense if you get a sweetheart deal and are willing to deal with Azure's crap, and I'd never recommend anyone to use anything Google any more.

[1] https://www.flyingpenguin.com/?p=54722


Ping is one of the oldest players in the business, they were founded in 2002 and had one of the earliest identity PaaS in the market (at least as far back as 2012). Haven't used their products much though.


Ping Identity run the largest enterprise identity platforms on the planet after merging with ForgeRock last year. Think HSBC, JP Morgan Chase-scale.


okta is not "active-active" in a multi-region sense, they run in a single active AWS single Region per-tenant. You can pay extra to have a faster failover in a region level failure scenario:

https://support.okta.com/help/s/article/overview-of-enhanced...


What is active-active?


Being live in more than one region at the same time


I haven't looked at Cloud Run pricing but running Kubernetes in the cloud is pretty cheap these days and my experience with solutions like Cloud Run in the past is that they end up becoming expensive.

Kubernetes can be as complex or as expensive as you'd like but it's also fairly possible to run a pretty bulletproof simple Kube cluster.


Maybe we misconfigured Kubernetes?

Here are my concerns:

With Kubernetes is that you need to pay for a few node just to keep it up, and then you need to pay for your nodes, no matter how much you use them.

Remember that Cloud Run charges based usage, so if a service sits unused for a while, which often happens in a heterogeneous microservices environment, you don't pay for it.

Also autoscaling is slow (Cloud Run autoscales really quickly, about as fast as your docker can be loaded and started, which for me is 1-3 seconds, where as I found Kubernetes auto-scales on the order of minutes) unless you over-provision, which is costly. This lets one scale to zero even without much of a hit.

I also ran into massive issues trying to get GPUs to work in Kubernetes - it was a driver nightmare that has wasted weeks of time collectively over the years. Whereas they are auto-provisioned properly on Cloud Run if you request them.

Lastly job systems on Kubernetes are a nightmare of configuration. The built-in scheduler cannot handle a lot of jobs but Argo also has its own issues if you actually try to use it. We've wasted weeks of effort on this. Cloud Run Tasks just skips this and is ultra fast too and handles scaling up to do a lot of jobs in such a simple fashion.

Honestly, managing Kubernetes is just overall a pain that has little benefit.

It is really hard to figure out what the benefits of Kubernetes is from my point of view. It has been a massive source of pain and costs and lost developer time.


It varies state by state in the US but the level of insurance you're required to buy for operating a vehicle is incredibly low. Generally, the level of insurance you must buy is dictated by the terms of financing because the only collateral the bank has is the vehicle itself.

If I were in the market for a CyberTruck (I'm not) I'd be thinking long and hard about what I'm getting myself into.


In my state, you can get out of the insurance requirement by holding $30k in escrow (or at least that's what it was about ten years ago; maybe they've increased it). I guess maybe it's meant to be indicative that you could get your hands on more cash if needed, but that's basically only enough to cover a fender bender. If there's injuries or a totalled vehicle, it's likely not sufficient.


If there's injury, $30k would only cover a day or two in the hospital, might not even cover ER expenses, as they'll be charged at list price not the "negotiated" rates insurance companies pay.

$30k is an absurdly small bond to post for self-insurance, though I'd believe that the regulations have not kept pace with reality.


The cash rate is similar to insurance negotiated rates. Typically 25-30% off automatically, 50-100% off with a financial aid application. Doesn’t change the inadequacy of $30k medical liability in a serious wreck, but don’t think you’re stuck with list price bills if you’re hit by an underinsured driver and have no health insurance.


Are you in California? The requirement is to either a bond $35k or to carry a liability policy that covers at least $30k for death and injury and $5k for property damage, matching the deposit requirement. That will increase to $75k in 2025, which still seems far too low.


For comparison EU mandates liability policy of 1.22M euro for property damage and up to 6.07M euro for injuries. https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:52...


I'm not. I'm surprised it's so low in CA.


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

Search: