I think this conflates setup cost with operating cost. The painful part of self-hosting was always to get it working: writing configs, reading docs, SSH ceremony. That's exactly the part that agents can help with a lot, so the perceived gap to Vercel shrinks a lot. But the problems that come later are still there, you just discover it later. DDoS absorption, zero-downtime rollbacks, cert rotations failing at 3am, someone else getting paged when they do. An agent can write my firewall rules, but it doesn't carry a pager.
So I'd reframe the thesis a bit: AI didn't destroy the moat, it moved it. Saving the customer setup time is no longer worth paying for, taking operational liability for the customer still is. SaaS that only offer the former are indeed in trouble, and honestly founders should verify that their product doesn't have this issue.
He doesn't really use the reliance argument. Whether dropping the pause was the right call is a separate question from whether people were allowed to rely on it. They were: the v1.0 language alluded to a much higher bar for changing commitments than they ended up applying, and Anthropic employees described the RSP as binding many times (Habryka says he heard it on more than a dozen occasions). Also, Hubinger's post announcing v3 says Anthropic is responsible for that impression. If you made vendor decisions based on that like the author did, the complaint is still valid, even if the change itself was the right decision. Both can be true at the same time, they were right to change it and you were right to feel misled.
I agree, his takes should not be dismissed lightly. I'm not sure about "demand is fixed" though. I feel like software demand has been declared saturated at least a few times.
Exactly, if we look at what projects are on-going now, look at Startups, they are practically solving all the same thing and most of them will be dead soon, we need to finally reach the era where tools to "zeroshot" anything becomes widespread to create new problems, but even by then, we will have an oversupply of tech workers, many will have to convert to a different field, many will not want to be paid based on callcenter type of work which is prompt-as-much-as-you-can, understandably.
It's quite hard to predict what will happen, but in a few years, I bet the unemployment rate of tech workers will be really high, we can just look at how many jobs are currently already replaceable but the owner of it is just lagging in the implementation of automation, it's probably already the large majority of tech jobs.
"fixed" is definitely incorrect but there's probably a ceiling on how fast the demand can grow, just because other bottlenecks will take over at some point.
Agreed. The limitations of human context window and communication bandwidth restrict the complexity of large-scale software.
LLM will have an extremely large context window and extremely high communication bandwidth in the future. Therefore, even more complex large-scale software will emerge.
I saw what changed syntactically. I meant I don't really understand what changed semantically. And whether there is any context to why the change was necessary.
I'm guessing the parent is wondering why this is noteworthy enough to be posted and discussed in this thread, and so if there's context they are missing
Wow, auto-installing any plugin onto live production sites should never be opt-out. For those managing client sites under approved-plugin policies and PCI compliance this isn't just a minor inconvenience. "We sent an email" is not an alternative to consent, and pointing people to manually uninstall across hundreds of sites is not a good way to handle the situation...
> That's rich coming from the company that tried to kill it
This post is written by three of the authors of the JPEG XL spec, implementors of the reference and rust implementations of libjxl, and...longtime google employees.
Mozilla's position from when Chrome first dropped it to September 2024 was "the benefits it provides are not significant enough on their own to justify the cost of adding another raster image format to the Web" [0], which they say is a "neutral" stance. Then like Chrome they only agreed to try it with jxl-rs [1], which is still their present stance. They are a complete passenger in this whole affair, like all other standards, where they basically just copy one side or the other (usually the more conservative side).
It's really bizarre to me that this is presented as "killing the standard". Is Apple also killing mechanical keyboards and hobby electronics development because they're the only ones who don't support Web USB or Web Serial? I strongly prefer having JXL and Web USB/Serial in my browser (FF for the last 20 years), but come on. If we don't like how much power browsers have in software distribution, then maybe software distribution outside of browsers should get fixed.
> Is Apple also killing mechanical keyboards and hobby electronics development
Obviously not, but they are killing as web standards the things they omit. At present for something to be standard it needs support from safari and chrome. That's just the current state of the world.
If tomorrow safari removed support for png that would effectively kill it as a web standard (assuming it didn't lead to mass revolt ofc).
For context, Google initially refused to merge JpegXL as a strategy play to promote AVIF, which was in use by other teams (i think Photos?). Internally, chrome engineers were supportive of jxl but were overridden by leadership.
I guess today’s post represents a change.
I don’t have any public evidence to support my claim, sorry. Take it or leave it
We (Google) built JPEG XL (together with Cloudinary). The main photography mode and the JPEG compatibility mode is from Google.
Chrome decided not to be an early adopter for good reasons that they have publicly documented, but that did nothing to JPEG XL. Particularly, it did not kill JPEG XL. Others, DNG, DICOM, PDF, EPUB, iOS, Safari, etc. integrated it early regardless.
Yeah, but they left out that Chrome removed their own support for JPEG XL saying no one in the industry was in favour of it despite everyone seeing it was the future screaming for it and building support for it into their own products.
Chrome's blink was the only major browser engine not supporting it and that prevented it from becoming a web standard and they refused to acknowledge they were wrong.
Chrome only backtracked once jpeg-xl was subsumed into the PDF standard because if Chrome did not support jpeg-xl, they would by extension also not be supporting pdf.
jpeg xl is also now used for the latest version of the DNG raw image format, and the iphone now encodes raw images as jpeg xl in DNG. It's so clearly the future for photography that Google is holding back. Apple surprisingly has been the first with full support everywhere in their OSs and in Safari.
I just wanted to add the decision for JXL inside DNG was well known even before it happened and Chrome still said no. Adobe and Intel along with plenty of other players in the industry screams this is so good they are all adding support. But still Google said no.
And to make matter worst the publish the worst comparison document and benchmarks for AVIF against JXL.
Sorry to hear that. I feel for you. Getting an SaaS off the ground is hard and there's no guarantee that it will work. I know nothing is ever that easy and circumstances are complex, but my advice would be not to focus on the SaaS. Not that you should abandon it, but it should probably not be your only hope. You can't force an SaaS to work out. Focus on your mental health and the relationship with your family. Focus on job search. Continue with your SaaS but limit the time and money you invest.
I know its never as easy as just do this, this and that and everything will work out fine. I wish you the best.
The Loureiro case doesn't explain things much better than a conspiracy theory does. At least, after a few minutes of research, I'm still wondering if someone could have manipulated the suspect into finally acting on a decades-long beef that does not appear to have been escalating steadily. Perhaps the hard part is why he shot the other people, but an angry suicidal person could do that, whether they were acting alone or triggered by someone else.
So I'd reframe the thesis a bit: AI didn't destroy the moat, it moved it. Saving the customer setup time is no longer worth paying for, taking operational liability for the customer still is. SaaS that only offer the former are indeed in trouble, and honestly founders should verify that their product doesn't have this issue.
reply