I am calling it now. LLM hosting is the new web hosting. You will have a market of hosting providers offering you access to LLM compatible hardware (the Hetzners of the LLM world) as well as virtualised LLM access (the Heroku of the LLM world). These will compete along pricing, ownership axes while frontier labs will compete mostly on performance, integration and ease of use (think Wordpress).
That's the only way I can see frontier labs charging high enough to sustain the cash flow needed to operate as racing to the bottom is not possible for them.
It is interesting to think whether this is another "Cambrian" era like the smartphone OSes when you had Symbian, Android, iOs, Windows Mobile and so many others competing.
I work at a very big company and they just pay azure and aws to host claude and co for them.
So the hyperscalers already won for now probably.
At the end of the day, you send a lot of personal data to these endpoints. If you already host everything through microsoft already, LLM hosting is then a no brainer.
Is deleting a letter after an LLM generated the article an insurmountable task? These quaint signals only screen out the lowest of effort slop writers. Better than absolutely nothing, but barely.
It does remind me of the time a chef told me when he puts lemon juice over a dish, he would intentionally not remove any seeds that went on it because it was a signal of quality. I wonder if future slop chefs will intentionally place seeds on dishes that came from a box...
This was inevitable. The better question is if AI related hardware costs drop after the AI bubble implodes, will Apple drop the prices? My answer is negative.
I think we might end up in a weirder situation: Apple _does_ drop their prices back down to current levels for the same quantity of ram, but ASP goes much higher, at least for the Pro tier buyers. My reasoning is that depending on how the benchmarks look, many of us may try to go big on ram on our next hardware purchases to run models locally as a way of hedging either model costs, or to ensure access.
The economic reality is that none of the social media websites would pass the test without a significant hit to their profits. If you want to see the long-term economic projections of a social media website without the younger market see Facebook.
It's like asking tobacco companies to reduce the toxicity and addiction of their products, inevitable collapse.
The reality is that the business model itself is inherently toxic.
Even when cigarettes were smoked by ~50% of the adult population, there were restrictions on sales to children. Why can we not accept the same for websites?
Yeah. Take Firefox choosing to create PDF.js to have a clean minimalist sandboxed PDF parser.
Chrome instead used an existing one that has been the source of dozens of vulnerabilities.
Or Firefox pulling in a ton of anti-fingerprinting measures from the Tor team. Not even worth talking about anti-fingerprinting as a serious consideration in Chrome.
Rust - a mozilla effort that resulted in code from servo being pulled into Firefox - chrome is headed that way too.
Even WASM was definitely a security improvement over NaCL, and Mozilla also led the way on Flash replacements in the day, making one of the first JS flash players (in the end, the solution was no more flash, but hey, at least they tried).
Font sanitisation - originally a mozilla security effort...
[edit] correction - I looked this up - I thought they used the chrome version, but they wrote their own sandboxing layer from scratch. On top of that they go beyond Chrome's measures with containers that isolate pretty much everything tracking-related if you use them.
https://blog.mozilla.org/security/2021/05/18/introducing-sit...
That's on the desktop. I don't know about the situation on Android, but my impression was the codebases are pretty similar these days.
Where did you get the idea there was no sandboxing?
Cite? I think the timeline has issues there. That predates the CEO controversies AFAIK.
They did ditch a lot of R&D as their userbase kept shrinking due to chrome growth. 'course this sort of thing keeps coming up - yeah, I do think their CEO is overpaid ... and? Solution is what. Kill firefox off completely, hand internet over to chrome? Basically, where is this point going?
In 2018, Baker received $2,458,350 in compensation from Mozilla. In 2020, after returning to the position of CEO, Baker's salary was more than $3 million. In 2021, her salary rose again to more than $5.5 million, and again to over $6.9 million in 2022. In August 2020, the Mozilla Corporation laid off approximately 250 employees due to shrinking revenues after laying off roughly 70 employees in January 2020. Baker stated this was due to the COVID-19 pandemic, despite revenue rising to record highs in 2019, and market share shrinking.
Yes, the (significant) salary increases happened well after the servo team was cut.
In 2020 when that happened she was at 3 million at a revenue of 466 million or 0.6% of revenue.
They laid off 320 people that year. If she had taken a salary of $0 they could have paid them each <$10k with that salary.
I don't think the salary was appropriate, but like a lot of these CEO compensation things, it's not going to make a huge difference to the final problem. Which was people switching to Chrome which google was pushing aggressively everywhere. ... and I guess purists here abandoning them for... Chrome? Again, no idea what the point is here. Mozilla has flaws, so screw 'em?
Those 2 browsers used a rendering engine developed by Google. It would not be wrong to consider them partial chromium reskins with all the technical dependency it entails.
I would guess because "so good" does not equate with 100% and presumably the user's needs fall in that 5%.
Linux has been usable for non proprietary software for decades now. The fact that people are refusing to jump ship even when Windows actively undermines them and itself speaks volumes of people's aversion (or inability) to switch OSes.
Oddly I’m like mostly using proprietary Windows software on my Linux machine these days (games).
I also think the AI era goes very far in eliminating those 5% problems. I have a mostly non-technical friend who set up an old laptop with Linux for the first time and he told me that he’d never have been able to do it on his own without AI. Anytime there’s an issue, his solution is just a quick question or copy/paste away.
There is a difference between trying something and performing a test whose results are meant to be representative of most setups. A lot of people (in tech surprisingly) who see themselves as practitioners of applied science, apply flawed methodologies and try to generalise the results.
Just because testing is fast, it does not mean that it is free or cheap. Plus time is money and verifying every LLM setup every koolaid dev is proclaiming on bluesky could easily be a full time job.
Digital systems don't necessarily deteriorate immediately after the causal factors. Like technical debt, issues grow unnoticed and become visible gradually.
That's the only way I can see frontier labs charging high enough to sustain the cash flow needed to operate as racing to the bottom is not possible for them.
It is interesting to think whether this is another "Cambrian" era like the smartphone OSes when you had Symbian, Android, iOs, Windows Mobile and so many others competing.
reply