there is a DW channel documentary about Transoceânica (Rio to Lima ) the longest bus route, which I just watched last week. It is 5 episodes but well made. https://www.youtube.com/watch?v=C_ODFlqURxY
I will speak from my experience. I have diabetes and I try to manage it well, with workout. But sometimes when the sugars are high for a while, I can feel it, the sadness, the hopelessness. It took me a while to understand that is high sugar levels and a mild form of depression. Now I will do some workout when I feel that and after a little workout, I can see how my mind also start to feel better. This is not a solution for everyone who is experiencing depression probably but might help some who are experiencing because of high sugar levels.
> But the reality is that 75% of the people on our engineering team lost their jobs here yesterday because of the brutal impact AI has had on our business.
Adam is simply trying to navigate this new reality, and he's being honest, so there's no need to criticize him.
And this is why AI coding will eventually degrade into a mess. Enjoy it while it lasts.
AI eats up users caring about $company which makes library, library degrades because nobody is paying, $company goes insolvent, library goes unmaintained and eventually defunct, AI still tries to use it.
Vibe coding with libraries is a fad that is destined to die.
Vibe coding your own libraries will result in million line codebases nobody understands.
Nothing about either is sustainable, it’s all optics and optics will come crashing down eventually.
AI is destined to destroy software industry, but not itself.
Software does not decay by itself (it's literally the whole point of using digital media over analog). Libraries do not "degrade". "Bit rot" is an illusion, a fictitious force like centrifugal force in Newtonian dynamics, representing changes that happen not to a program, but to everything else around it.
The current degree of churn in webshit ecosystem (whose anti-patterns are increasingly seeping in and infecting other software ecosystems) is not a natural state of things. Reducing churn won't kill existing software - on the contrary, it'll just let it continue to work without changes.
You’re mostly right, libraries thrive by adapting to their surroundings. Mostly.
But after just months of being unmaintained, even the best libraries start to rot away due to bugs and vulnerabilities going unfixed. Users, AI included, will start applying workarounds and mitigations, and the rot spreads to the applications (or libraries) they maintain.
Unmaintained software is entropy, and entropy is infectious. Eventually, entire ecosystems will succumb to it, even if some life forms continue living in the hazardous wasteland.
I struggle to fully grasp everything you postulate. Please help me understand.
Your original point was that libraries do not need companies behind them. From what you have written here a reason for that is that (web) libraries mostly create churn by introducing constant changes. What I think you follow from that, is that those libraries aren't necessary and that "freezing" everything would not do any harm to the state of web development but would do good by decreasing churn of constantly updating to the newest state.
What I struggle to understand is (1) how does AI fit into this? And (2) Why do you think there is so much development happening in that space creating all the churn you mention? At this point in time all of this development is still mostly created by humans which are likely paid for what they do. Who pays them and why?
Bit rot isn’t some mystical decay, it’s dependency drift: APIs change, platforms evolve, security assumptions expire, build chains break. Software survives because people continuously adapt it to a moving substrate.
Reducing churn is good. Pretending maintenance disappears is fantasy. Software doesn’t decay in isolation, it decays relative to everything it depends on. And it sounds like you don’t know anything about Newtonian dynamics either.
That is one take and certainly possible and negative but I think people create libraries for different reasons.
There are people who will use AI (out of their own pocket for trivial costs) to build a library and maintain it simply out of the passion, ego, and perhaps some technical clout.
That's the same with OSS libraries in-general. Some are maintained at-cost, others are run like a business where the founders try to break even.
It's just interesting because most of the talk is programmers talking about AI taking their job by replacing them not taking their job because it's taking away revenue from the business.
Reminds me of the problem with Google & their rich results which wiped out and continues to wipe out blogs who rely on people actually visiting their site vs. getting the information they seek without leaving Google.
I expect a lot of business disruption because of AI. Agree it's not the same as employee replacement, but it adds to the sort of fog of war around what effect AI is really having.
Anything open source will be turned against its authors and against ICs.
We thought it would give us freedom, but all of the advantage will accrue to the hyperscalers.
If we don't build open source infra that is owned by everyone, we'll be owned by industrial giants and left with a thin crust that is barely ours. (This seems like such a far-fetched "Kumbaya, My Lord" type of wishful thinking, that it's a joke that I'm even suggesting this is possible.)
Tech is about to cease being ours.
I really like AI models, but I hate monopolies. Especially ones that treat us like cattle and depopulate the last vestiges of ownership and public commons.
it's a real shame no one warned us this would happen when a bunch of corporatists and opportunists wrested the term "open source" from the advocates of true freedom in the late '90s.
Also the FSF squandered its opportunity being RMS’ hobby / support organization and skipped a lot of important discussions, even before the skeevy behavior they’d been ignoring came to light. I used to donate in the 90s but … really feels like that was just flushing cash.
ChatGPT came into the picture long after the open source issues we’re talking about were apparent. AI companies are making it even worse but solid advocacy in the 2010s or 2000s would’ve been helpful.
I'm just not sure how to connect this rhetoric to the facts of the source link, where a hobbyist attempted to extend some source-available code to support a new technology, and the CEO of the for-profit company who owns the license said he's not allowed to for business reasons.
You can be and I am sympathetic towards the CEO! I wouldn't accept a PR for cannibalize_my_revenue.txt either. But if we insist on analyzing the issue according to the categories you're describing, it seems undeniable that the CEO is a corporatist, and that he put an unfree license on his repository to stop people from freely modifying or redistributing it.
There were more-or-less two original spheres of OSS. There were the academics who were too "pure" and holier than thou for everyone else, and then there were commercial FOSS that OS'ed because something already reached its reasonable lifetime potential and it was cool to give away the plans to a cult classic to let it live on in some other mostly permanent, mostly released form. When OSS becomes a mindless pattern, an absolute prerequisite to investment, and/or ceases to be released without regret, resentment, and/or strings attached, then it's not cool anymore and becomes toxic.
There's no such thing. Even if on paper "everyone" has an ownership share, in practice it's going to be a relatively small number of people who actually exercise all the functions of ownership. The idea that "everyone" can somehow collectively "own" anything is a pipe dream. Ownership in practice is control--whoever controls it owns it. "Everyone" can't control anything.
> I really like AI models, but I hate monopolies. Especially ones that treat us like cattle and depopulate the last vestiges of ownership and public commons.
I would dispute whether the tech giants are "monopolies", since there's still competition between them, but that's a minor point. I agree with you that they treat individual coders like cattle--but that's because they can: because, from their standpoint, individual coders are commodities. And if automated tools, including AI models, are cheaper commodities that, from their standpoint, can do the same job, that's what they'll use. And if the end result is that whatever they're selling as end products becomes cheaper for the same functionality, then economically speaking, that's an improvement--we as coders might not like it, but we as customers are better off because things we want are cheaper.
So I'm not sure it's a consistent position to "really like AI models" but also not want the tech giants to treat you like cattle. The two things go together.
> we as customers are better off because things we want are cheaper
Why privilege that side of the equation over "we as workers"? Being a customer isn't all there is to life. I happen to spend quite a bit more time working than shopping.
It's not a matter of "privilege". It's simple economics: if the same functionality can be provided more cheaply, that's a gain to everyone. The gain to customers is the most obvious gain, and it's what I focused on in my previous post--but it's also a gain to producers, because it frees up resources to produce other things of value. But the producers have to be willing to change how they make use of resources in order to take advantage of those opportunities.
> I happen to spend quite a bit more time working than shopping.
Then you should be a lot more worried about AI providing the same functionality you were providing as a coder, but more cheaply--because that makes you, or at least you as a coder providing that functionality, a commodity that's no longer worth its cost. So if you want to avoid being commoditized and treated like cattle, you have to change what you produce to something that AI can't do more cheaply than you can.
Stop enabling corporations' theft and exploitation.
Don't FOSS by default, unionize, embrace solidarity, and form worker-owned co-ops that aren't run by craven/unrealistic/non-business founders if you want any sort of stability.
IMO, the only ethical and legal way to build LLMs on the entire output of all human creativity, that still respects rights and won't lead to feudalism, is conforming to the actual legal requirements of fair use that are being ignored.
According to fair use doctrine, research models would be okay. Models used in education would be okay. Models used for public betterment by the government would be okay, etc
Pie in the sky version would be that models, their output and the infrastructure they run on would be held in a public trust for everyone's benefit. They wouldn't exist without consuming all of the public's intellectual and creative labor and property, therefore they should belong to the public, for the public.
> Tech is about to cease being ours.
On the hardware side, it's bad, as well. Remote attestation is here, and the frog is just about boiled when it comes to the idea of a somewhat open and compatible PC as the platform for general computing.
It was kinda cool while it lasted, glad I got to see the early internet, but it wasn't worth it to basically sign away for my great grandchildren to be peasants or belong to some rich kid's harem.
It does give us freedom. In fact, it arguably gives more people freedom, as non-programmers can create now simple tools to help themselves. I really don't see any way that it reduces our freedom.
They commoditized their complement to their hardware/infra, that being software. Good for them and the value of tech will shift to what is still scarce relatively.
Some of the critics in the thread are… odious. I’ve written down some of the GH handles, because if I’m ever hiring again, I wanna make sure I’d never hire some of these folks.
I don’t understand how someone can display such contempt towards the maintainer of a thing they’ve used for free.
> I’ve written down some of the GH handles, because if I’m ever hiring again, I wanna make sure I’d never hire some of these folks.
You can block accounts on GitHub and add a note as to why. Might be simpler and more accessible later on than a random TXT (plus, it probably updates if they change their username).
Note that blocking also means they can’t contribute to your repos. Which you may not care about anyway.
Thank you, that’s indeed much cleverer. Unfortunately I’ve closed my account this year, trying to put my money where my mouth is and not furthering the goals of GitHub or Microsoft.
What's wrong with microsoft and github? I can't lie, between this comment and the "writing down people's github handles out of spite just in case" you are coming off as someone with a lot of grievances.
Understandable, but I’m definitely a lot less bitter than it seems.
Well, Microsoft is vile. I won’t expand because there’s plenty online on the topic. And I don’t like their acquisition of GitHub, which has turned into an ecosystem for laundering open-source code through LLMs.
"Sorry, we cannot give you the job because even though you're qualified and passed our interviews, you were such a meano to Adam! That is a no-go at this organization"
Half the people in that thread have this mentality that just using tailwind is enough contribution, so therefore GiVe mY oPuS MoRe InFo
I thought we learned years ago that exposure doesn't keep the lights on. That mentality is nothing but entitlement
One comment stated that "it's not our fault the founder was unable to manage his finances to pay his people" well if open source worked the way people try to act like it does, he shouldn't have to pay anyone, right? But here we are
So you're going to make a list and ensure they don't find employment? That's really an overreaction. Most people say "personality fit" is a thing, but I feel you're downright disgusting if you're tracking people like this and actively ensuring they don't get a chance because of a difference of opinion. It's vindictive and in poor taste and not objective at all.
Nice, nothing like a little personal bias to inject into an interview process. If you can't handle criticism and you're just looking for sycophants, you're probably not the type of employer or hiring manager most people want to work for anyway.
Fair enough, I tend to avoid overly negative people. Criticism can come from a good place to suggest improvements (radical candor), but agreed that some of the comments are just personal attacks. I think we're both in agreement that those aren't people we'd want to work with.
You can use a product and still be critical, especially when layoffs happen, truth is there are a lot of things we don't know about their finances – tailwind definitely is successful by any metrics, they have corporate sponsors that alone give them a healthy MRR (I count at least $100k/month from the sponsors page alone)
I sympathise that it sucks having to fire people, been there. But it sucks more to get fired.
I am one of those critics, but I never used Tailwind. A layoff of that magnitude is horrific, but if what they are describing as their business model is true, they really really need to rethink it. I wonder what the size of their marketing team is like, and if they were involved in the layoffs. Seems like they need some help there. I found the "downvote" spam in that thread, for reasonable posts, to be quite off-putting, and that led me to my remarks.
Specifically thanks to equity takeover. I’m human, so yes, I can be prejudiced. People who succumb to mob mentality to hate F/LOSS maintainers fall under such prejudice.
I don’t want to sound harsh and didn’t mean to offend you.
I am not 100% sure about that - I usually find AI written CSS to be slightly visually flawed and almost always logically flawed.
The way you write websites that actually work imo, is you understand how your chosen CSS layout engine works roughly, and try to avoid switching between layout modes - traditional to flexbox to grid to flexbox again down the tree can drive the most brillant devs utterly mad .
But seriously, after a certain complexity threshold, it becomes impossible to tell what's going on and why.
And if you don't think about it in advance, it's very easy to reach that threshold, especially if you don't get to write the whole page from scratch, but have to build on the work of others.
AI (and many frontend devs) do write-only CSS - they add classes until the code they write looks right.
But code like that tends to fall apart under multiple resolutions, browsers, screen sizes, devices etc.
I am not a frontend dev, and came pretty late to the frontend party. That said I felt that anything that obscures the raw CSS makes it much harder to deliver UI that works right, as it peppers hidden side effects across your code.
That's why I wasn't too keen on CSS frameworks like Tailwind - I found that when writing frontend code the writing part takes up the minority of the time, it's producing a well thought out layout flow is what is actually the biggest sink of time and effort.
That said, I'm not a frontend dev, and I'm to too good at CSS - but not horrible either - so I defer to the judgement of others who are pros at this, its just my opinion and experience.
If you want a bunch of tailwind class slop, then yes. Otherwise, A lot of context engineering is needed if you want it to write modular tailwind components properly for large projects where consistency is important.
> Otherwise, A lot of context engineering is needed if you want
I am not seeing that. I have a few AI-assisted projects using tailwind and scrolling through it now 99% of it looks... completely modern and professional. I had previously asked it to "completely refactor, a rewrite if needed, all the tailwind/css/app styles. ensure visual and code consistency across pages".
Modern coding tools add tons of their own content, but none of the above was "a lot of context engineering".
Isn't that an article about using a frontend aesthetics prompt in order to avoid the AI tells? A lot of the with-aesthetics pages look pretty good imo.
It's describing the problem and also giving a solution. The problem of vibe coded sites all looking the same is very real however, if you don't consciously and actively guide the LLM towards being different, as described in the article.
And design too. I shouldn’t be able to tell Claude designed your site/app, but it is too often the case. Good taste still remains an advantage thankfully.
"You should learn how to vibecode and ship whatever works enough, as fast as possible, to get bought for a wildly disconnected from fundamentals valuation." This may sound flippant, or low quality, but it I assure that it is not intended to be. It is derived from observations of the current tech macro. Quality does not appear to matter, ethics do not appear to matter, sustainability and engineering rigor do not appear to matter; it appears that all that matters is "Start up. Cash in. Sell out. Bro down."
I would love to be proven wrong, truly, because this is a path to the death of craftsmanship, deep knowledge, and to some extent, curiosity, in the domain.
It satisfies the dream of a business with no people. As Doctorow illustrates it, like plugging the Fisher-Price steering wheel into the drive train of the business.
> Quality does not appear to matter, ethics do not appear to matter, sustainability and engineering rigor do not appear to matter
I don't know why people keep saying this, as if quality, ethics and sustainability mattered before and every developer was a pure artisan of their craft. In reality, having been in many companies and looking at their codebases, it has always been slop, with very few exceptions.
Yeah, no kidding. I was alive 20 years ago, this isn't like talking about the 1800s, what exactly was different with the craftsmanship and ethics back then?
You might be right but even then this feels fundamentally really immoral
The sell out is the biggest fundamental issue in this equation because it is the part of the equation which doesn't reward Quality,ethics,sustainability and engineering rigor overall.
Welcome to the AI bubble fueling it.
I genuinely don't know but I think AI prototyping/using it for personal use cases are fine but when we completely start to vibecode, if your project is complex enough, you will reach problems and all the other factors/researches point out. In my opinion, for longevity, vibecoding is not the deal.
But as you said, longevity isnt rewarded. I really hate how the system has become of just selling businesses.
I feel like as such the businesses who are truly passionate about their product (because they faced the problems themselves or are heavily interested in it/passionate about it) might win "long term"
To me trust feels the biggest resource in this day and age. Information era has now been sloppified. Trust is what matters now.
I don't know but I will take the slow but overall steady route. There is a sense of commitment with human trust which I feel would set apart businesses and I will try to create side projects with that initiative
One of the ways I feel like acheiving it while still getting the shipfast aspect is that I just build things for myself, vibe coding in this case can help and I launch it for public, if there is interest in any product or smth, I will try to respond and try to add feedbacks fast (perhaps still using vibecoding) but in long term, I try to promise to keep the code lean (usually approx 2-3k lines of code at max) and then if I see prospect and interest about the idea, I have tried to think that a middle way is either rewriting or completely understanding AI generated code to its core and having a very restrictive AI access afterwards any product feels good and then the trust aspect of things can be gained.
I don't know too much about side hustles. I just build things for myself in whatever I want mostly I must admit using vibe code and end up usually sharing it online/deploying it for others as well if it might help.
That’s unproven, but suppose it’s true: what’s your alternative? If we are in fact facing widespread unemployment, what’s going to be better than UBI at avoiding societal collapse? Billionaires paying private armies to contain poor people is a straight-up sci-fi dystopia but even that depends on enough people having money to buy things from their companies.
If we truly hit the point where we have more people than jobs. That we hit AI improving at miraculous paces that we cant even reskill people. I think it would be better to essentially have make work programs. Have basic qualification programs where you are guaranteed a job. People need a purpose. Throw every person capable of getting an engineering or science degree into labs. Massively expand teaching, nurseing and medicine so there is extremely personal care just by the sheer numbers.
This is so fucking dumb. I hate when software engineers try to solve problems. You are good at one thing, do that.
The rest of us will struggle without your help because that's what we been doing. We are literally struggling to fulfill our purposes because we have jobs.
Then use it to pay for services like healthcare and education so that everyone has a safety net and opportunity to thrive without just giving everyone enough cash so that they are incentivized to slack.
DDT has been banned, cigarettes are all but banned, leaded fuel has been banned. Nuclear energy has been banned in Germany.
The industry wanted all of that and did not get its way after some time. You can ban "AI", make companies respect copyright. You can do all sorts of things.
Since "AI" can only plagiarize, countries that do the above will have an edge (I'm not talking about military applications that can still be allowed or should be regulated like in treaties for nuclear weapons).
I disagree, the only real issue with UBI is the amount of inflation it will cause. Germany has something nearly approaching UBI and they are doing fine.
Did you read the expanse? The earth of the expanse is full of crime and destitution. People apply in the tens of thousands for every lottery slot of school or jobs. People just wallow in nothingness. The people fleet earth for mars and the belt just to have a basic sense of purpose.
If we are to just have UBI. Have basic sustenance for no effort, while we have unlimited entertainment and porn at our finger tips. It would be a disaster. I would literally we rather have make work programs.
Agreed, it's one of the only ways forward I can think of while still maintaining markets in some part of the economy...that is, if you care about the human condition at all. Plenty of these tech leaders seem to want to replace humanity though, so this will be an uphill battle.
What is your alternative, when the price humans can sell their labor at dips below what is necessary for them to survive? All these takes about "UBI will demolish the human spirit" or whatever are just ridiculous when the alternative is "starve to death".
Just doing nothing isn't great for the "human spirit", but UBI doesn't mean people can't find their own goals to pursue. The idea of something where people are not longer required to work to survive is hard to accept since many people haven't seriously considered how they could meaning outside of their careers
counterpoint : my father had realistic expectations for what he wanted to do post-retirement.
what actually happened was that he sat around purposeless because it turns out that the motivation of producing a paycheck or product was actually the reason he did things. He stopped showering, became depressed, and neglected his health.
And this isn't an uncommon reaction to the open-ended 'free-form' life post-retirement. Some people very realistically need to have some level of structure imposed on their life or otherwise be taught how to create that structure themselves. I think this will be a very real problem whenever UBI gets closer to reality.
I see two alternatives, one that people find new ways to do productive work with or in the presence of LLM, or massive social unrest, rebellion, war and/or starving to death, followed by a reset. I.e. the way human nature has responded to similar imbalances in the past.
Funded by an automation tax as proposed by Martin Ford. Not holding my breath on either count. We mustn't upset the 1,000 or so billionaires in this country in any way for they are wise and they are kind and only bad things will happen if we do.
But chin up, peasant, each and every one of us can dream of one day being a billionaire as well if only we act as wise and as kind as they do.
> But chin up, peasant, each and every one of us can dream of one day being a billionaire as well if only we act as wise and as kind as they do.
(I know this was written satirically) but this is a nice example of doublespeak and I immediately got reminded of it.
I wouldn't say that we have reached 1984 level, there is still some decentralization where you can get hosting and then self host from small vps providers as well etc.
Not that most people do such things tho. Internet is still heavily centralized but overall, there are still outlets of escape legally and you are able to sometimes even talk to vps provider owners themselves directly in some cases if they are small enough.
But still, each year although we get away from 1984 the year, we get near to 1984 the book.
As much as I am pro AI and I really am very pro AI, there is definitely an emperor's new AGI vibe amongst the tech bro and billionaire classes. I can only attribute it to a compulsive need to oversell everything and then deliver 25 to 50%, a state everyone is so used to now that if you try to be honest and make claims that state what you can really deliver, they will assume you can only deliver 25 to 50% of what you are claiming and therefore the guy promising twice as much gets the funding.
This makes me happy that I'm nearing retirement but that switch flipping is being delayed by my hourly rate going up for possessing forgotten knowledge. Sigh...
Yes, I have no idea who's this magical "we" in your "We can simply". To me this seems like a textbook coordination problem leading to a tragedy of the commons- even if you got 99.9% of the world into your "we", the remaining "defectors" would have a massive benefit from using AI to replace human labor.
No, I have the same question as that other poster. It is not a bad faith question.
There are a lot of problems that would be solved immediately if "we" (i.e. all of humanity, or all of the U.S. or some other country) decided collectively to do something: climate change, nuclear weapons proliferation, war, and so on. But that's effectively wishing for magic -- there is no way to get everyone to collectively agree on something, so unless you explain how to cope with that fact, you haven't actually made any progress.
Given that I personally don't control humanity as a hive mind, what can I do to fix this problem? You haven't proposed an answer to that.
the strong interpretation is that you mean we gotta do something. and it's really not "simply" even because "we" needs to include everyone and whoever is a renegade will get more benefit.
so if "say" is an euphemism for "do" it seems an obvious question what exactly do we "do". that's another reason why it's not "simply". even if everybody was ready to do something as one, if you think everybody just knows what we should do because it's so obvious you'r mistaken.
sure it's asked a bit sarcastic but sarcasm isn't banned right?
Not only can we not just do that (you did not even define what you mean), but China is coming out with models that are good enough for this purpose - and they are, because they are open, everywhere.
Indeed we need to revolt against AI and force every other big powerful nation to do the same thing. Yet unfortunately that seems like a big joke until AI has destroyed their society too.
The reality of UBI in the United States is that it's going to go from being something freely given to being something that is a full time job to maintain, and then it will be cut or replaced with services that are specifically designed to be as cheap as possible. Until we're all living in terrafoam, birth-controlled and warehoused until we die.
Is it that mass unemployment will lead to caring more for one's family again, resulting in proper family structures that take care for their elders like in the past? I hope so.
When you talk in meaningless terms like "traditional workers" and "tech bros", all it tells me is that you have divided the world into people you like and people you dislike and mourn / celebrate accordingly.
If ones position for "other people" was "they should pull themselves up by their bootstraps" then the same applies. If your position was we should stop/slow/consider the march of progress - well you lost to 30 years of moving fast and breaking things.
I suggest and ask for nothing but consistency, irrespective of if you like or dislike the people who are affected.
> It is "progress" when class traitors displace labor, but it is "heartbreaking" when a class traitor gets displaced by other class traitors.
it becomes clear that the original comment was a pointless strawman of a position that nobody holds. A class traitor wouldn't be expressing sympathy about displacement in the first place. It only seemed to make sense because, when you say "tech bro", people superimpose the general category of technologists who think they can make the world better on top of one specific stereotypical guy who believes all the worst things they've ever heard a technologist say.
unfortunately, it doesn't seems like tech bro gets displaced by other tech bros at all and more like corporates running costly ephemeral branding as tech bro by abusing other tech bros works.
What's the difference between tech bros and corporates? Isn't being a tech bro almost by definition about getting to the point where your can sell out your company and your principles?
Well, I never read the artcicle because paywall, but there is a WSJ headline today about a $160k mechanic job at Ford that can't be fulfilled because no labor
I don't buy it. They failed to build a sustainable business model and are now suffering the consequences. Everybody is leaning into AI because it works (in the sense that it pays the bills). Saying the layoffs were because of AI offloads the blame.
Just posting the "75%" without context is a bit of an odd choice. He explains why in the podcast, but it still feels like he should have specified immediately to avoid assumptions about scale.
What? He could have said 3 if he wanted, but he wanted it to sound worse so he said 75. I know its inferrable how many people it is, but if the guy laying them off doesn't care to say the number, why should someone else when posting this?
Both of those numbers in isolation dont tell the whole story. Saying firing 3 people sounds like a wednesday at a big company. Saying firing 75% of the staff indicates the impact that those changes will have on everything about the company. The latter is more useful.
oh, come the fuck on. it's "AI made us do it" drivel that companies began to justify layoffs with in 2023 (!!!).
Tailwind is just another FOTM frontend thing. I saw dozens of them come, gain some popularity, then abruptly disappear once the marketing budget ran out.
He mentions that tailwind is more popular than ever before but their revenue is down 80% so unless he’s lying about that it makes sense rather than tailwind going out of style.
However, why is that even surprising? Tailwind is essentially a frontend css stylesheet. What business could there possibly be around that?
I understand, they have UI kits, books, etc. but just fundamentally, it was never going to be easy to monetize around that long term, with or without AI.
Tailwind also has a compiler of sorts (so you only include in the bundle the exact styles you need) and a bunch of tooling built around it. In an alternate universe it could have been a fully paid enterprise tool, but then it might not have caught on.
The comment you are responding to said their revenue is down 80%. So they did monetize training and services, and I don't see how that would have been a problem long term if AI didn't come along and make all of that unnecessary.
Yes. The point I was trying to make was that after the initial hype disappears, sales in those categories would probably taper off regardless. But it is purely my opinion.
I come from India. Many years ago I remember I was having a tough day and very less money on me and sitting by a busy road side tea stall along and sipping my Rupee 2 tea , looking at passers by and traffic. I don't remember anything more but that thought often comes to me randomly and makes me happy even after about 20 years. Life brings joy at unexpected corners.
This is certainly alarming for US auto manufacturers. Tesla is the only successful EV car company which is able to somewhat compete with BYD, but for many it is hardly an option because of it's leadership.
That's to say nothing of the cost. Assuming there were no extraordinary tarrifs on China/BYD, the entry-level offering would be in the $10K range which is about 1/4 the cost of a base Tesla Model 3.
Thanks for sharing. I am particularly interested in knowing your experience with ConnectRPC. I wasn't aware about this library. I use Go and JS implementation in one of our webapps and it was painful specially on the web but once working it was beautiful. Just curious if you feel it is a much better developer experience with ConnectRPC ?
In any case, my experience so far has been great. The connect-es generator takes your protobuf and generates a TypeScript library which looks pretty good. You can use ConnectRPC both from the browser and from Node.js (the latter can leverage HTTP/2, while the former uses the native fetch() or a function you provide).
I am trying to make a constructive feedback and not just critical if I sound that way by anychance. I spent a bit of time but it's hard to get the product. Instead of the team photo on home page, you could show some images of what you mean by the product personalization. Honestly people don't have much time to read through and understand a product , which has a simple value proposition.
Thanks this looks intersting and I am going it to try it later. I have old Axiom 49 and it really doesn't work that much with modern DAW as it is assumed it's old and outdated. But I like the form factor and it is solid. I hope I can make it work witht his one ?
Thanks that was a great breakup of cost. I just assumed before that it was the same pricing. The pricing probably comes from the confidence and the buzz around Gemini 3.0 as one of the best performing models. But competetion is hot in the area and it's not too far where we get similar performing models for cheaper price.
I love SQLite and this is in no way I'm making a point devaluing SQLite, Author's method is excellent approach to get analytical speed out of SQLite. But I am loving DuckDB for similar analytical workloads as it is built for such tasks. DuckDB also reads from single file, like SQLite and DuckDB process large data sets at extreme speeds. I work on my macbook m2 and I have been dealing with about 20 million records and it works fast, very fast.
Loading data into DuckDB is super easy, I was surprised :
SELECT
avg(sale_price),
count(DISTINCT customer_id)
FROM '/my-data-lake/sales/2024/*.json';
and you can also load into a JSON type column and can use postgres type syntax
col->>'$.key'
duckdb is super fast for analytic tasks, especially when u use it with visual eda tool like pygwalker. it allows u handles millions of data visuals and eda in seconds.
but i would say, comparing duckdb and sqlite is a little bit unfair, i would still use sqlite to build system in most of cases, but duckdb only for analytic. you can hardly make a smooth deployment if you apps contains duckdb on a lot of platform
depending on the size and needs of distributed system or application im kind of really excited about postgres + pg_lake. postgres has blown my mind at how well it does concurrent writes at least for the types of things i build/support for my org, the pg_lake extension then adds the ability to.. honestly work like a datalake style analytics engine. it intuitively switches whether or not the transaction goes down the normal query path or it uses duckdb which brings giga-aggregation type queries to massive datasets.
someone should smush sqlite+duckdb together and do that kind of switching depending on query type
It's not an index, it's just (probably parallel) file reads
That being said, it would be trivial to tweak the above script into two steps, one reading data into a DuckDB database table, and the second one reading from that table.
reply