lemarchr

It was so refreshing to see a video with a clear and direct title. In 2026 this would have been "We cracked open a motor. You won't believe what we found inside! <emoji>". Checking the channel, it seems they have succumbed to this for their latest videos.

That would be a different experiment.

Some here are arguing that mechanisms used by LLM providers already derail the goal of "the very best, most precise words at every single decision point", therefore the author is misguided.

The author has expressed a preference. Assume that there is a sequence of tokens, such that it is considered the absolute best by the author. This particular method of watermarking makes it less likely to generate that sequence, by definition.

I feel their argument would have been clearer and stronger if they had spent more time exploring the alternatives, and whether these alternatives would be just as effective. It is trivially easy to remove invisible tokens.

Like it or not, there is a public good to being able to identify AI generated content, and a small degredation in quality is tolerable in my opinion.

I don't think anybody has to worry about this issue though. Manual writing, coding, and proof reading continues to be an option. Where AI output is nothing to be ashamed of, the tools are available. For everyone else, there will be LLM providers that ignore EU law.

I also suspect those who have decided to leave their employer will be more likely to create, maintain, and even keyword stuff the skills listed on their LinkedIn profile. This would be a poor proxy for real world skill level and their impact on the departed organisation.

And despite the click bait "billions wasted" headline, at no point does the article offer any information on the alleged billions wasted.

It's global news to me. Hello from Australia!

Edit: Ah wait, I think I get what you mean. By "global news" you mean it's been deemed worthy of sharing to an international audience by media outlets, as opposed to a crowdsourcing news source like HN. Is that right?

Argument from authority is a pernicious fallacy, and typically effective too. You were right to call it out. I must admit I overlooked the sarcasm, however.

The comments section gives me intense astroturfing vibes, especially the way it abruptly starts and stops between 2014.

Edit: Although the negative comments stop too. Maybe people were just more passionate about yoghurt a decade ago.

> In a practical world this is an outrageously useless platitude.

As an industry we went from treating engagement as an art form, to treating it as an optimisation problem. Along the way we got very, very good at it, at scale. While profitable, we now know that the methods used to drive this engagement are harmful. Many countries recognise this harm and have introduced legislation to limit or ban methods used on vulnerable groups, such as children.

It is interesting that we, as an industry, do not talk more about the harm we are complicit in causing. Somewhere along the way we normalised and accepted the idea that addicting features are desirable, and that we are not responsible for the consequences.

With all said, an industry is not an individual. You and I may care about this problem, but it is not clear how to fix it. At the very least, as individuals it would be good to avoid contributing to addicting features as a matter of principle wherever the opportunity arises, lest we become the equivalent of digital drug dealers.

While the analogy they offered may not be a good fit, I agree with the sentiment. My takeaway was not that addicts have zero agency or ability to help themselves, but that it's unethical to rely on this as the sole mechanism for minimising harm.

See https://www.datastax.com/our-people/jonathan-ellis

Wikipedia lists them as a founder. Perhaps their author bio is outdated, or Wikipedia is. Not sure about your friend.

Not the original commenter, but I have often been asked by non-technical colleagues to answer questions about a spreadsheet they've sourced from someone/somewhere else. "So-and-so shared this spreadsheet with me. Can you please take a look at let me know if ...". Sometimes this is delegation, but frequently comes down to a lack of skill.

I don't understand why there's so much negativity here. From my cursory perusal of the docs this looks like a simplified, vendor agnostic re-imaging of something like CDK, with cool tooling including visualisations and out-of-the-box support for local dev. Where's the beef?

You think cloud is too expensive or unnecessary? Fair enough, this tool is not for you.

You think cloud infra is necessarily complex because you need to support <insert use case here>. You're right! This tool is not for you (yet?).

You don't need this because you already know <CDK / Terraform / whatever abstraction is already in your repertoire>? I agree, the juice is probably not worth the squeeze to learn yet another tool.

Are you approaching cloud for the first time or have been managing existing simple infra (buckets, queues, lambdas) via ClickOps and want to explore a feature constrained (hence easy to grok) Infrastructure as Code solution? Maybe give this a look.

While it's still early days, I suspect there will be many who will find this useful, and congratulate the authors for their efforts!

My thoughts exactly.

We don’t do LeetCode—our interviews are like regular dev work. Candidates get access to an existing codebase on Github complete with a DB, server, and client. Environments are Dockerized, and every interview's setup is boiled down to a single "make" command (DB init, migration, seed, server, client, tunnelling, etc)

This is what all employers should be doing and the part that Litebulb should really focus their marketing on. Skills based assessments that mimic real-world responsibilities are far more predictive of a successful hiring outcome. If you want to pass LeetCode style interviews, you practice LeetCode style problems. Have you then proved that you can do the job? Sure, if the job is to solve LeetCode challenges; however, it's far more likely your role will involve adding a feature to an existing codebase while keeping all the tests passing, including appropriate test coverage, ensuring your solution is clean and maintainable, etc. Designing a problem set like this is hard, and there is overhead in setting the project up and maintaining it over time. I wish I had the time to design an interview like this.

Without knowing all the details or having gone through the experience myself, I can imagine it being positive for both employers and candidates. I assume a lot of the backlash I'm seeing is an allergic response from previous exposure to other tools that focus more on the automation part to the detriment of candidate experience, such as those selfish, impersonal and awkward as hell asynchronous video interviews.

I'd be willing to give this a chance.

There's a GitHub link to source in the top-right corner, if you're interested: https://github.com/feross/magickeyboard.io

Are they capable of fulfilling the role of primary learning material? Or are they supplementary to more traditional texts?

It must be that time of the year again: https://hn.algolia.com/?q=Papers+We+Love

Edit: Not meant as snark, just a playful observation of how often this is shared. Remains a worthy contribution to this day.

> Private college prep secondary schools (at the very least—often it's private schools all the way) on the other hand are overwhelmingly the norm in that set.

To which data set are you referring? Data from 2019 found that 80% of Fortune 100 CEOs hold undergraduate degrees from public institutions[0].

[0]: https://www.forbes.com/sites/kimberlywhitler/2019/09/07/a-ne...

Malware on the user's machine could catch the password in either case.

This is frequently still beneficial for the organisation, even when you're doing it for the wrong reasons, and is almost a necessary stepping stone towards achieving true understanding. It's hard to identify nuance until you discover some strongly held belief doesn't apply as well you once thought.

But I suppose if you could go back and do it differently, you would, so I can appreciate why you consider this to be a mistake.

Why do you consider this to be a mistake? I've thought about this in the past, but can't really articulate my motivations, and in the end I haven't followed through. Some of my previous colleagues have fit into the gap between co-worker and friend. I would never spend time with them outside of work and since we no longer work together I'm not sure what I'm really missing, so I'm curious to hear your thoughts.

Great overview with a minor nit-pick; it's Cross-Origin Resource Sharing rather than Request Sharing. It describes a server's willingness to share its resources across origins. The client's request isn't the thing being shared.

For those using Java, it's also possible to achieve this using Testcontainers[0]. It's nice to have the test DB setup co-located with your test code, and if you're using JUnit it's really easy to manage and align the DB lifecycle, e.g. fresh DB for each test, shared across class, etc.

----------

0: https://www.testcontainers.org/modules/databases

Ah right, you did say that! My fault for not reading the article more carefully!

If you're developing apps in containers, there are platforms that make it really simple to deploy. For example, Cloud Run (Google), Fargate (AWS), or Heroku. What the parent comment is suggesting is that building a platform like Cloud Run, Fargate, or Heroku is much easier on Kubernetes. Kelsey Hightower (principal engineer for Google Cloud) put it well when he stated, 'Kubernetes is a platform for building platforms.'¹

You can still deploy apps directly onto Kubernetes and it works very well for this purpose, but it will require a lot more learning than one of the platforms listed above. If you enjoy learning, Kubernetes is an incredibly powerful and satisfying tool to have in your kit, and the initial learning curve isn't as steep as some make it out to be. If your goal is to deploy apps as quickly and simply as possible however, go with one of the pre-existing platforms.

If you still want to learn Kubernetes then a really great book is Kubernetes Up and Running. It goes into just enough detail at the right point in time to make it simple while still being useful. If you do a bit of Googling, you might find a free copy of the book that used to be offered by Microsoft to promote their Azure Kubernetes Service. Otherwise there's Kubernetes the Hard Way² but that's more focused on administering the Kubernetes cluster itself, rather than how to use the cluster to deploy apps. You'd need a pretty convincing reason to administer your own cluster rather than spinning up a managed cluster on GKE or EKS.

My advice: - Grab a copy of Kubernetes Up and Running - Install minikube on your local PC - Experiment and have fun learning

Hope this helps.

---

1. https://twitter.com/kelseyhightower/status/93525292372179353...

2. https://github.com/kelseyhightower/kubernetes-the-hard-way

It's probably fine, but reading about a single Postgres container in a Kubernetes cluster with backups to S3 gives me sweaty palms. I hope the author has fully tested their disaster recovery plan.

Also worth noting that it's a lot less resource hungry than the official Kubernetes distro. Kubernetes documentation recommends 2 CPUs and 2GB of RAM per node just to have enough left over to actually run any workloads.

k3s is specifically designed for resource constrained environments (e.g. Edge, IoT, CI/CD) and can get away with as little as 0.05 CPU and 256M RAM (on worker nodes).

I don't recommend this as a way to save money on infrastructure but it does open the door to further use cases.

> It made me the the confident woman I am today ...

> nemiah posted to Indie Women on March 5, 2021

Bills don't just magically go away if you refuse to pay them.