Man! I’m disappointed this article didn’t mention the coolest .plan hack of all time: you could link your .plan file to a named socket so whenever someone fingered you, an arbitrary program would run that spat out whatever output it wanted back to the client. Made for some very fun hacks, like randomly generated text every time you were fingered, or showing things like the current load on the server, etc…
This article resonated tremendously with me, speaking as someone who has been coding since the 1980s and had to develop my own taste the hard way, through lots and lots of mistakes. Whenever I see a software demo built by someone mostly using agents to do the work, I have to question whether it's any good on the inside - or whether any real intuition or judgment went into its design. On the flip side, maybe none of that matters anymore - as long as it works, who cares how it was built.
For majority of people, including business executives it's the latter. No one cares about artisan code. Software is just a means to an end for them and not an art form.
People treat their computers like any other appliance. They don't care if it's riddled with bugs, they'll gladly restart their computer or refresh the page at infinitum to get it to work.
I think it's also why marketing and designers get more of a say in the software than engineers. UI polish matters more since it increases sales.
This is the hardest thing to come into terms with in this industry.
I'm quite sad that I have to disagree with you and agree with GP: if quality siftware mattered to people, Microsoft would have gone out of business a long time ago.
If quality software mattered, javascript wouldn't be the most popular way to write desktop software.
If users cared at all about quality software, we wouldn't have 800MB chat apps.
Of course yes, and it's even easier in some cases. Since there's a language that defines the functionality (e.g: formal methods) that the agent can use/veriy against
Yes but that 10% is all that matter to the people that matter to the business. Like a car, most people care about how it looks and how they interact with it... the only people who care what is under the hood, for the most part, are other engineers.
I just want to add that programmers have only relatively recently become so abundant (i.e. replaceable) that they have so little say in how software should be built.
Programmers are still the experts in the room, but they don't get what they ask for for free anymore.
It resonated with me too, and even as a once fairly prolific coder that now uses agents to do all the coding, I feel like this is all relevant and does matter. The agent does whatever you ask and provides thin judgements based on what it thinks is your motivation and context, even feeding it the context it will often make bad generic choices. I think this is the part hard to optimize because it's not a verifiable reward, it's the feel of the software. I think Herdr is a great example, I've tried several dozen terminal multiplexers / session aggregators and they are all vibed and different, but this one felt so much better to use because of its choices and actually increased the number of parallel sessions I was able to keep going. My own projects I find I have to stress every detail, that is what makes the software good now is really diving into it, the experience feels like just as much of a craft as coding was and 'taste' captures the concept for me.
For what it is worth, I appreciate you sharing your experience. Looks like you have more programming experience than I have in life, so your comment matters to me. It also gets rather close to something I wanted to evaluate in the post: "whether any real intuition or judgment went into [its] design". I think that's a big question to dwell on these days.
Ironically, though, I had written a post that had people debating whether I put any intuition or judgement into. Would've been real ironic to instead add a post-scriptum that said "Yep, none of it matters. I just slopped something in 5 minutes and it easily went to the frontpage. Look how easy it is!" but unfortunately that is not why I write, not what I intended to achieve, and not something I am equipped to do. I care about my work, which is why I care about your comment as well.
“as long as it works” is doing a lot of work here. Did the people building it say “no, again” enough times, and at the right times, in order to make something that _actually_ works instead of _merely_ works?
People who care about their health aren't ignorant about that. I think that can apply to software too: if you care about safety or privacy you might think twice before using vibe-coded software.
That is a significant leap in logic. A vibecoding tool is not a McDouble. Nor is Chris Kempczinski the same person as Dario Amodei.
The biggest difference is that there are only so many ways to eat a McDouble, but you can actually wield a vibe coding tool with skill. There's nothing stopping you from looking at the code and asking for structural reworks or refactors that please your code taste. In fact one of the best use cases for AI coding tools is vibe refactoring old apps where the code is straight up horrible.
> but you can actually wield a vibe coding tool with skill. There's nothing stopping you from looking at the code
It’s hard to give something like “vibe coding” a rigid definition, but I think we can all agree that it’s inversely correlated with "looking at the code". The more I examine the code while building, the less the activity is about vibes.
So, used this way, it's true that it's not a McDouble anymore, but it's also not so much of a vibecoding tool anymore either.
i have never had a "mcdonald's mcdouble", whatever that is. i cook my own burgers because i am a human being and i want to eat real food, not nauseating slop that should only be fed to livestock
The LLM has its own intuition and judgement which at this point is quite good. I’m not just throwing that around, I use LLMs 12-14 hours a day and for a long time reviewed atleast 75% of the code.
I've spent a day fixing a feature my very junior colleague pushed, and which was merged by my should-have-known-better colleagues who have almost as much experience as me.
We have access to the same models. Why did her version suck and my features don't need fixing?
These look great. The page says "Thousands of designers, developers, and content creators use HugeIcons Pro". Is this accurate? The project seems to have only about 493 downloads a week on NPM, and 194 GitHub stars. I would have expected a lot more given the number of people you claim are using the library.
(Edit: For clarification, the NPM stats refer to the number of times the library was fetched from NPM's servers, not the number of times a developer incorporated it into a project. For popular NPM libraries, this number can be in the many millions of downloads per week. My confusion stems from the fact a library used by thousands of projects will have substantially higher numbers in NPM than what is seen here. It is of course possible that this number counts developers incorporating these icons not through the NPM library.)
This looks cool! Your website really should feature a couple of sample videos, though, since without finding the links in your post here it is hard to tell what the final results look like.
I'm a co-founder at Fixie, a Seattle-based AI startup. We're launching an experiment to showcase what's possible with Fixie's real-time LLM-powered voice platform. You can check it out at https://hisanta.ai - you can have live voice chats with Santa, Mrs. Claus, Rudolph, and other friends (toggle the naughty/nice switch at the top for some other characters!)
We'd love to get your help testing this and get any feedback -- especially from people with kids :-)
This is not just a thin wrapper around LLMs. To get the performance to this level, we've had to do a lot of engineering to wire together ASR, LLMs, and TTS models efficiently, and are doing some clever things to hide latency. Happy to answer any questions about the technology.
The Fixie platform (https://fixie.ai) powers the app and embeds the logic for managing voice sessions with LLM agents, which can be built either through a no-code UI or using our AI.JSX (https://ai-jsx.com) framework for building LLM apps. We support a range of ASR and TTS models under the hood as well.
I’m no fan of the WEI proposal, but the headline here is inaccurate. Rick expressed a belief that criminals were amongst those voicing concern over the proposal, not claiming that all opponents are criminals, as the headline suggests.
That's a distinction without a difference in my mind. Claiming that criminals are against the proposal is "poisoning the well", even if he's not accusing _all critics_ of being criminals.
> my suspicion that there is significant intimidation from criminals who are afraid this feature will disrupt their illegal and/or unethical businesses, and I don't give in to criminals or bullies
The notion that this is poisoning the well is also supported by the general pattern: retreating on an idea that others thought was bad, avoiding literally any discussion of the technical/cultural criticism (despite that it apparently still existed legitimately enough to postpone the proposal for now) and 2/3rd or more of the post spent implying that [all|some|who knows] of the opponents were abusive bullies, and that maybe the idea was actually fine. (Setting the stage to absolutely re-introduce with the new framing "inauthentic traffic" in 6 months.)
Note also, avoiding any discussion of the web open or how it entered into the decision to postpone, supposedly. Though, that also wasn't said, just sort of implied with the half-hearted smiley-faced plea for compromising.
I am not emotionally invested because I see this as inevitable and don't care, I'll go live in the woods soon anyway, but this is just icky to read.
I think what's happening here is that he's just too deep in to google's corporate culture and can't imagine normal people perceiving google as a threat. His first paragraph is basically this,
"Thank you for your comments. We are not going to listen to your comments. We're surprised that not listening to your comments has led to people trying to contact through other means. The only reason I can think of to object to WEI is crime so enough of you are probably criminals that I'm going to mention it in the first paragraph as an accusation."
I have deployed similar-looking devices in places like volcanoes in Ecuador. On each of these boxes we put a prominent message in Spanish and English: "This is sensitive research equipment, please do not touch" with contact information.
OP here. Thanks for the kind words. I certainly didn't feel like this was an easy raise, but then again it's the only time I've done this and my comparison points were other first-time founders who raised 2x what we did with less than we had done. Yes, these are all pretty senior, well-established folks, not kids straight out of college.
The main point of my article was the surprise around the extent of the VC network and the helpful interactions with them. Before going through this process, VC was a black box to me. Now, a lot less so.
A few folks below have pointed out that I must have had an extensive VC network to draw on. Not quite. I knew 3-4 VCs casually from having worked at a couple of other startups. None of them invested in us by the way. It certainly didn't hurt to get their advice. Now, of course, I have a rolodex full of dozens I could potentially call up at some point, which is useful. I can see how founders who have done it before likely find it a lot easier to get the ball (and the checks) rolling.
OctoML | Multiple Positions | Remote, Seattle, or Bay Area | Full-time | https://www.octoml.ai/
OctoML is changing how developers optimize and deploy machine learning models. We are the creators of Apache TVM, a popular open-source compiler that transforms ML models into highly-efficient binary code optimized for the specific hardware and model architecture. We are also building the Octomizer, a cloud service that enables developers to optimize and package their ML models through a modern web app as well as a rich API surface.
Series C, over $130M of funding, company is 100 people and growing to 200-250 next year.
Long-time former Googler here, author of go/l6talk (internal presentation on getting promoted to L6). While there I coached a lot of people on the L5-to-L6 promotion process. It seems a lot of Googlers get disillusioned about getting promoted above L5. Broadly speaking Google's culture is heavily focused on levels and promotions, which is too bad, since many L5's are doing just great and getting paid extremely well.
If it weren't for the fact that Google makes level information public within the company, I doubt there would be quite so much emphasis on getting a promotion. Most people want to get L6 for the recognition and the fact that it elevates you above other engineers in terms of status and influence. When I worked at Apple, levels were private, so you had to treat everyone the same -- a person's influence was not dictated by what level they had on their company profile (there is no company profile at Apple). In my opinion this approach was much better, and significantly lessened the internal competition for promotions.
