tabs_or_spaces

Kind of interesting how people feel this way. What if I said that anthropic's current valuation was dependent on a bunch of folks getting rejected? Because if you were hired, then others like you would have been hired. Would there have been as much progress or luck with many folks like you? Don't take me wrong, it's not that you are bad at what you do, it's just that you can't guarantee the same present outcome at all.

It's the same claim like bitcoin. Many would say that if they bought bitcoin, then they would've had a massive return. But the bitcoin price is the way that it is because of people not buying early. There's no guarantee that the price would've moved the same if you bought earlier.

Figured I'd just give some food for thought to hopefully make you feel better.

I did something similar, but with current youtube search results and not really stress processing on HN data. My technique was simply saving my search terms and topics, and then putting that under a crop and continually updating the results. Super simple, but resulted in such a better experience for YouTube videos for me.

I call it "re-algorithming" your data. Basically you're not happy with the algorithm of a product, then you use it's data and introduce your own.

It also seems like folks are tired of the youtube algorithm and want something different instead? I'm willing to bet that many sites and apps have this opportunity in them.

I think the concept is amazing, so don't take the rest of my comment too seriously.

But why do we anchor naming to someone who made something popular? Why do we upvote similar names too? For example, openclaw became popular then there were similar projects with the "claw" naming.

It's just a very fascinating psychological thing to do, and I don't understand why. Because if you copy part of a name, generally you'll never get more popular than the original name. So why do this?

I'm completely fine with it. Currently I work with a set of principles that I try to uphold when working with agents (many of which are existing principles that tries to scale for the volume of code being written).

* Code is meant for reading, not writing. You will spend more time reading code so optimise for reading. I got this advice from a previous Principal Engineer and it has stuck with me ever since. Yes you can generate slop, but you have to be able to read it. Otherwise, for me, sloppy code becomes the context for the next feature, then the feature after that, until the codebase is really messy and brittle.

* The other principle I work with is that more code means more bugs and more recently security vulnerabilities. We're churning more and more code, and looking less at the code. Another agent can very easily find vulnerabilities in your website or app. Sometimes I feel that this is inevitable with the way folks are doing things

* Then my most recent principle. Always add value, know your customer, know your product, know your business metrics. I'm seeing more and more the statistic of "I merged X PRs in a day/week". Just like before AI, these things are very superficial. How many tokens did it cost? How many dollars did it cost? How much revenue did it earn? How many customers are adopting the feature? This applies to internal and customer facing features. The amount of vanity metrics that AI can create now is astounding, and people are still falling for it regularly.

* In a new project, I start with a few commits to see if the agent understands my desired architecture. Once that passes, everything else falls into place and slowly trust the code more and more. Sometimes when I need to add a complex feature, I would sit more with the design than the implementation, then I would skim the generated code to see if it's adjusting accordingly.

This is about consolidation of power.

Open weight models are catching up to the frontier. It also seems like frontier models reached some limit, whether this is capex related, business model related or something else. Nobody knows but it's happening to all frontier labs it seems.

It's been fear mongered many times that ai will kill us all. But this time, a person with around 2-3 months of tenure at anthropic managed to go viral with no previous social media account activity, gets picked up by all news outlets and kicks off yet another round of fear mongering.

So the real questions to ask

* is all of this to increase the valuations before IPO?

* what is the real barrier to entry for open weight models to be used by the public?

* what is the Financials of these companies showing that's causing this outcry on safety?

People need to think really critically about the things happening around them. Don't only just look at the face value of what's being presented here.

It's really disappointing to see social media dismissing gemini so easily.

I think the worst thing we can do is have loyalty towards models. I used to be loyal towards Claude, and my viewpoint changed dramatically when I used codex.

I highly recommend that if you are someone who only used one model so far, that you really give another model a shot and see how it goes. It's very eye opening and gives you a more holistic perspective.

Vendor locking is a big problem when it comes to models, and I hope the software world doesn't do this blindly.

It feels like people are comparing Uber the brand in their own country and applying that to other countries

* People assume that Uber is a big brand in Nigeria and Uganda, this might not be true

* People assume that Uber is popular in these countries, this might also not be true

I started using llm's on a $20 claude license, then it became a $100 license.

Then all the mythos warnings happened, then the fable access issues started.

At that point, I just wanted to try something else because I didn't want to feel vendor locked by anthropic.

Then I tried codex on the $100 plan. I've never been happier. It does everything claude can do, without any token limit issues or anything for me.

I don't see myself going back to claude anytime soon. I also highly recommend folks to not vendor lock yourself to a model. You need to try multiple models for yourself and see what's possible. There should be zero loyalty to any frontier lab because of how fast things change and how volatile model access, uptime and customer experience can be with these things.

> LLMs aren’t lazy. They don’t cut corners because a simpler solution feels good enough. If they know how to solve something thoroughly, they will.

This is a severe misunderstanding in how LLMs work..

I don't know how this got on my front page....

I'm more interested in the business impact of this

So you spend billions of dollars training the model, only for it to be used in the US.

Then interesting to see where most of anthropic revenue comes from. If it's the US then they're fine but if it's global then they'll see a drop in revenue?

Then add to this decision, companies are going to significantly reduce their token spend.

So what does all of this mean for their IPO?

Super appreciate you replying to my comment.

I think I understand where you're coming from now. What confused me is that the post is written in a way that it seemed like what Fable was doing was actually better. Maybe I should've looked at post as an exploratory post on Fable instead.

> How can a LLM be assigned an emotion as being "proactive"

I can't edit my post, this is wrong. "Proactive" is defined as a behaviour instead of an emotion.

Thanks to everyone pointing it out!

How can a LLM be assigned an emotion as being "proactive". This is highly misleading to anyone that scans just the headlines.

What actually happened is that the user started a prompt, and Claude took $12 worth of tokens to resolve the issue. How it did so was basically looping until it got to the answer

How is this proactive? It's literally being token greedy and maximising revenue for the LLM owner. People really need to be putting on business hats at this stage, because we are being lead to believe that "more tokens = better". It is not, there are efficient ways to solve a problem and there are inefficient ways to do so too.

Each problem solved incurs a cost, and is expected to yield an ROI at some point. This is how we should be viewing things now.

For me it depends on who you listen to

If you're following a bunch of people who are from LLM labs, you're going to be more incentivised to tokenmaxx because it's in the Lab's best interest tonget you to behave that way.

Practically, many companies aren't labs with endless runway. Companies hopefully follow a PnL model. And when you look at things with that lens, many of the times the LLM use case falls apart.

You're seeing a bunch of companies starting to realise that tokenmaxing yields very little ROI.

Even the LLM labs, the guy that spent $1+mil tokens has nothing to show for it in terms of revenue to the company. And you have to keep sinking that much into AI for ... "features".

There are some good use cases for AI. I ended up with a positive ROI on a greenfield project myself, albeit on a small scale.

The way that AI has been making people have totally irrational decisions on executive, pure business and technical standpoints is simply mindblowing. I don't understand how people can't take a step back and see what's actually happening from a macro perspective.

it's a net positive for robinhood, not the trader necessarily

Product market fit and profitability are two different things.

Arguably product market fit was fond last November already. I don't think agents were the turning point that caused this.

Profitability, not yet. For me, it depends on whether companies are seeing a positive ROI from their ai investments. This website is skewed more towards big tech companies, but everyone that's not big tech needs to see positive ROI with using ai tools. Short term might see a profit, but medium to long term we still need to wait a bit.

So the article doesn't mention how the individual should prepare, but rather that government should prepare. Are individuals just that powerless against these perceived outcomes with AI?

My opinion is that AI can be a force for good, but why is everything so aggressively framed as a class war? Why must such a path be taken?

I wonder if this is how things were during the industrial revolution as well

Why would you replace an existing codebase like this instead of forking the repo instead and then making the changes?

Why is there only a fine and not also the seizure/forfeiture of the stolen property and the derivative works/products built on it?

Like the fine means nothing to meta, and they'll still be the beneficiaries of their infringement.

In this current state, you really just need to have enough money to bypass this lawsuit and be on your way.

There's probably a few things to consider

* How much CPU/token usage does openclaw users use in general? Similarly, how much does high volume openclaw users use vs "normal" claude high volume users?

* Are there political elements we can't see that's affecting this? OpenClaw and anthropic doesn't have a good history in general and this is just a continuation of that?

Something I don't understand, there's a lot of complaints yet people are reluctant to stop using the service? Are folks already vendor locked or is it a case of "well, this doesn't seem to affect me?" The consumer behaviour of these complaints is very interesting.

> When I went through tough breakups? I lost myself in open source... on GitHub. During college at 4 AM when everyone is passed out? Let me get one commit in. During my honeymoon while my wife is still asleep? Yeah, GitHub. It's where I've historically been happiest and wanted to be.

I've never had such an obsession to a platform or an activity as this. Some might say this is unhealthy, but I admire folks who can reach this level of obsession in their craft. It's just a joy to read about for me

It looks like it's this person's fault?

* you can't blame ai if your production token is on the same machine as the staging/ development environment?

* you can't blame ai if you didn't know that the production api token gave access to all apis.

Like if this is the level of operational thinking going into this app, then I'm sorry no ai agent or platform can prevent this from happening.

Everything else in this "post mortem" is performative at best.

The only real question one could ask railway is why do they have api endpoints that can affect production available? Maybe these should only be performed on the platform itself instead?

Once I wrote the perfect piece of software. It was so perfect that there was literally no bugs for months.

How could this have happened? Well, the code was shipped but no customer was running it in production.

I find HN to be a bad resource to ask for learning resources. I previously asked for help in learning how claude works but no responses.

Maybe one pointer for others is that people are genuinely curious about learning new things, but as experts we choose not to engage these types of posts, why is that?

OP, in your case you need to move from theory into actually building your own agent and make it do things. Start by solving small problems and then make them more complex over time.

The headline seems to be flashy indeed, but ai didn't really solve this imo.

They just seemed to fix their technology choices and got the benefits.

There's existing golang versions of jsonata, so this could have been achieved with those libraries too in theory. There's nothing written about why the existing libraries aren't good enough and why a new one needed to be written. Usually you need to do some due diligence in this area, but no mentions of it in this post

In order to measure the real efficiency, gnata should've been benchmarked against the existing golang libraries. For all we know, the ai implementation is much slower.

The benchmarks in the blog are also weird. The measurement is done within the app, but you're meant to measure the calls within the library itself (e.g calling the js version in its isolated benchmark vs go version in its isolated benchmark). So you don't actually know what the actual performance of the ai written version is?

The only benefit, again, is that they fixed their existing bad technology choice, and based on what is observed, with a lesser bad technology choice. Then it's layered with clickbait marketing titles for others to read.

I'll probably need to expect more of these types of posts in the future.

I think Sora was technically impressive as a concept. The way it was managed as a product wasn't good.

There didn't seem to be any marketing for it. Like I can't even remember an ad for it or any content creator type of person pushing Sora actively.

To get access to Sora I believe you needed to be on a paid plan?

It's really difficult to get user generated content going when it's behind a paywall.

It's also hard to tell if this means that openai is in trouble, or if this is just a badly managed product that deserved to be killed. With the negative sentiment on openai, folks might think the former.

I would add that getting customers, especially paid customers for your app is not easily solved with ai too.

Only a few get lucky with funding, only a few have a profitable business.

It's very interesting to see the opinions in the answers vs the off site opinions.

On the site itself, the opinion seems to be that stackoverflow is not dead and it's okay if there's reduced traffic and questions to the site.

Yet off site, the opinion is that stackoverflow is dead and added to that, the community isn't very friendly anymore.

It's just really interesting to see the perspectives between the two.

> In the last 60 days I have written over 600,000 lines of production code — 35% tests — and I am doing 10,000 to 20,000 usable lines of code per day as a part-time part of my day while doing all my duties as CEO of YC.

LOC will never be a good metric of software engineering. Why do we keep accepting this?

I can generate 1 million LOC if I really wanted to.

As long as LOC is the main metric for these setups, they will never be successful.

I think the advice is good but maybe the title could be improved.

> But for Engineer A’s work, there’s almost nothing to say. “Implemented feature X.” Three words.

To me, this is the main problem. Engineer A is unable to describe the impact of their work, how the work affected the business. Your manager isn't responsible for promoting your own work, you are.

> Engineer B’s work practically writes itself into a promotion packet: “Designed and implemented a scalable event-driven architecture, introduced a reusable abstraction layer adopted by multiple teams, and built a configuration framework enabling future extensibility.” That practically screams Staff+.

Maybe it's just the narrow of the article, but if promotion only looks at complexity and not quality of delivery and impact on the business then this isn't a good engineering team to be in.

There are many cases where simplicity is celebrated and recognized. It's up to the engineer to know what the impact of their work is, if they can't do that then that's on them.