Of course, computers now are so expensive no one can afford them anymore
Well of course everyone has to switch to cloud, even for non AI use.
Try and buy a new server nowadays.
Another great result of this AI craze.
787 publicly visible posts • joined 15 Sep 2009
NOO!!! You don't understand!!!
This needs a GPU (which is super expensive), is made for agentic AI (which is expensive and useless), and has this thing that state is maintaned on the server and somehow sent to the client, like, you know, IBM 5250 terminals from 1977.
We are well over the "reinventing the wheel" phase. We are reinventing a square wheel made of stone, to be used in a boat instead of on a road. Expensive, proprietary, utterly complex, and totally not needed because cheap, free solutions already exist, which are screen, tmux, mosh, and tssh.
But we need it for agentic AI. We absolutely need it. (not)
I'm a freelance consultant. I always push for open source and on-prem solutions, but still most if not all SMBs are all-in on MS or Google or AWS. And since these vendors say that they are GDPR and NIS2 compliant, there is absolutely no push to change to something different. Only cash strapped micro businesses consider open source locally hosted solutions, and only for one reason: they are cheaper.
I think this is inevitable if you use such a method. Also a simple copy/paste from a user wanting to keep some text from the page will fail.
I am using Anubis web firewall to try and stop scrapers. Of course it might not work, but at least a human can still copy/paste my pages if they like.
More and more web sites that I visit are implementing some sort of challenge or puzzle that I must solve to prove that I'm not a bot from a US AI company before it will let me in in order to keep their servers from being ground to dust by the AI scrapers
The issue here is more complicated that that: "legit" AI companies are scraping the hell out of everything, but at least they are doing it with a user agent that states who they are, so you can block them if you like. Then there are search engines turned AI companies (Google, Bing, Meta, and some others) that still use a proper user agent string, and these you can block, too... but if you block them you disappear from Google or Bing, so you are committing web suicide. They are the purest evil because you cannot keep Google from accessing your site if you want visibility, but you cannot tell them "index my site in the search engine but do not use it for AI training", so you know you cannot win. And some of them (Meta) are pounding hard, with a lot of requests per minute, killing sites that have database-intensive page generation.
Then there is the great majority of "under cover" AI scrapers (usually Chinese, but not only Chinese) that use a random / fake user agent. And while I absolutely want to block them all (not because they are Chinese, but because they are hiding their name and scope) I cannot do it without some trickery like Anubis web firewall or similar systems. And this course of action is mostly useless because they will soon learn how to trick Anubis into letting them in.
Yes, it's nuts, but still every business and even a lot of government entities rely completely on MS, Google, Amazon. And it seems that no one sees the need to actually own and control their IT assets (hardware, software, etc) and their data that is stored on such systems.
I'm a strange beast and while I'm Italian, I use a US keyboard and a Denmark language, Denmark time format and day/month names and Italian numbers and currency on my Linux Mint. So dates are in the right format dd/mm/aaaa but language is English, week starts on Monday (Italian style) currency is Euro, decimal is comma, number separator is dot. And Betterbird 140 ESR works and the dates are right.
Or it shows that the AI can now improve itself, which is the whole point of the system. Once the AI can improve itself, it will become super intelligent in a very short time, Bezos will fire every worker in every sector and then sell his products to nobody because nobody will have the money to buy anything. But it will probably be irrelevant because we will all be dead by that point.
3 newbie mistakes:
1- have a "normal" email address to receive important notifications. I have one reserved for this, WITH NO SPAM FILTERS, and it's secret so it does not get spam.
2- have the recovery email on the same domain that might need to be recovered (same service provider, etc). It must be a totally different one. My admin contact for OVH is not hosted on OVH, this is the idea.
3- have only one MFA token (app, generator, whatever) so if it fails you are out. Have multiple accounts or multiple MFA auth tokens for the same account or keep the seed somewhere so you can easily reload it into a new app / device. Or have single use codes on file, or whatever.
Yes, open source software can help here, and the technolgy for everyday computing (not AI slop) is here since forever.
Still EVERYONE USES MICROSOFT.
Why?
Well, because everyone else uses Microsoft. It's that simple. We must break away from this idiotic loop, but it's not easy at all, even less easy since there are tons of rules, policies, etc, and "best practices" that have grown over time around only one service, and that's Microsoft ecosystem. No one will break away from it unless required by law.
The issue is you pay for AI even if you don't use it.
- hardware costs are 10X, and you need hardware even for "classic" computing.
- energy cost is raising fast, and you need energy even for "pen and paper" computing, if you want heat, light, or air conditioning.
Because of these two facts, even grocery cost is increasing because the people who make, transport, and sell groceries need computers and power. Everyone needs computing and power nowadays.
Also, because of the crazy expense in AI from idiotic IT companies, their products (even without AI integration for the customer) will cost you a lot more. And since everything a subscription, the cost increase will happen NOW and not "if and when we buy the new version of software X"
It is not as if internet use is a zero-sum game: every time a bot scrapes a page, a human is denied it
Well, it turns out that it indeed is. I manage some services that are quite niche like storiaememoriadibologna.it, a web site dedicated to information about people and event and places about Bologna, Italy.
This site has limited resources and the queries tend to be very database intensive. Human visitors are quite sparse, this is not a high volume site. Well, now AI scrapers make for 95% of the traffic, and they tend to kill the server, generating outages that last from minutes to days.
I also manage a medium sized e-commerce for a sports apparel reseller. This site has 24 cores, all flash storage, 128 GB of RAM and it has quite a lot of human visitors.... well, AI bots can kill it anyway, quite easily. If the site can generate a page in 0,2 seconds but the AI bot makes 20 request a second, it's easy to see what will happen. And again, humans are denied the access to the service. And the owner of the site is not happy to pay for a monster server that is capable to serve 5 times the usual human traffic, but chokes on AI slop generators coming to scrape everything.
This is exactly what's happening, but there is probably even more. While everything in this article is true (and I suspect the percentages are far worse than this), there is also another issue. LLM output has been actively tweaked to always look like a competent, well researched, proper answer. An LLM will never tell you "I'm not sure" or "I don't have enough data to give you a decent answer". The meme where the person asks the AI if this mushroom is edible and then gets poisoned and the AI says "sorry, I was wrong" is spot on. LLMs are tweaked to produce an output that looks good, not that is good. And it works, because people fall for it.
And this is the least worse part of the tweaking. Then comes the steering of the political or ideological output towards the direction that the current techno oligarchs want. Goebbels' wet dream.
And then comes the capability of AI to collect data and infer results (right or wrong, it does not matter), Palantir-style. Stalin's wet dream.
Are we happy? Of coruse we are. Or at least it seems that everyone I know is happy about this.
The issue here is that a lot of common people are unintentionally investing in AI because banks are selling aggregate funds (sorry I don't know how they are called in English) that contain AI stocks to their customers. And rest assured that once the crash comes in sight, the banks will promptly "sell the debit" to unknowing / inexperienced customers and we will all lose money.
Because politicians are part of the system.
Lobbyist fund politicians who get elected because they have the money to campaign for their election.
And the real issue here is that this (unless it comes down to bribes and corruption) is perfectly legal.
And while Europe is living (right now) its worst heat wave (WORST FOR NOW, not worst overall) it's also green-lighting a ton of useless AI data centres.
I'm not American so I don't know how their pension works, but here in Italy if you have insurances and such (a sort of private pension fund) you usually don't get to choose where they do invest money. They invest in other funds that invest in other funds, etc. Who knows where does the money actually go in the end?