Dear literally everyone writing, "This is what's wrong with Google Search" articles:

Your articles literally read like, "This is what's wrong with repeatedly slamming my member in a car door."

Just stop, already. You do not have to use google search. For anything. Ever.

It's been nerfed to the point that DDG is 99% as good, which you can even use with JS turned off, and customize it to do things like remove their AI search suggestions and stuff like that.

<voice actor="William Shatner" mode="maximum scenery chewing">
So... just... stop... using... Google search. It's... that easy. See?
</voice>

Thank you.

@rl_dane Probably all written by people who weren’t alive before google search came along so do not realise there used to be others - and there can be again.

@rl_dane well said!

Also, courteous mention that DDG has https://noai.duckduckgo.com - No AI, no customisations required. Increase traffic to this URL and let them know we don't want AI!

@nigel

Totally. I remember when I'd look at people really funny if they weren't using Google search.

Now it's the reverse. XD

@rl_dane yes, same.

Sigh. Times, they change.

@rl_dane

Point of order, am I, as a search engine developer, allowed to write about problems with Google Search and what I'm doing differently in my engine?

@amin

Obviously, yes, and anyone is allowed to write such an article if the punch line is, "So why the heck are you still using Google, hoser?!?" ;)

@paul @rl_dane dear DDG. Can we swap the URLs so users don't have to. Showing the desire for AI is so low that it's not worth burning the earth for it. 🔥🔥🥵

@justin @paul @rl_dane Well I mean that's the reason why they don't isn't it?

@sotolf @paul @rl_dane probably. Hopefully they lose plenty of money for this stupid decision.

@justin @paul @rl_dane Well I would guess they will when the bubble finally pops, I just hope they will get out of fashion, I have at least one podcast that I'm listening to because one of the hosts obviously uses ChatGPT to do their essays, and it's annoying since it's one that I've listened to for over a decade. The whole lists of 3 things, generalised boring language and whole shebang..

@sotolf @justin @paul @rl_dane My hopes with you🙏
Because I'm also listening one podcast (Radio-T) for near 15 years and ≈1-2 years ago they started not to discuss various IT news, and blogposts, and stories from life, but almost only "AI" news (2/3 of podcast time now dedicated to the news about different LLMs and how they replace programmers soon )

@justin @paul

I kinda wish there was a mode that said, "Show me cached, pre-computed 'AI' suggestions for search terms, but do not compute any new ones for me."

@sotolf @justin @paul

Dang, that's horrible.

I heard MSFT is planning a datacenter that would use the equivalent energy as 1/6 of the entire U.K.

Even if they used their own fusion reactor, that. is. not. ok.

@sotolf @paul @rl_dane AI as it is today can die a fiery death. One day we'll use it for good. But today really isn't it.

@justin @paul @rl_dane I'm not sure if LLMs will ever be "good enough" for anything, machine learning sure, there are a lot of places that can be helpful, but I don't think generative AI is here to stay.

@sotolf @justin @rl_dane
Agree.
Generative AI has the problem of not being reliable and the way I see it the only possible endpoint is the realization that the amount of time it takes to spec out and describe all possible edge-cases to cater for specific needs, avoiding avoid hallucinations and factual errors is much more wasteful in time and money than paying a human developer/researcher/author etc.

It needs to go away now.

@paul @justin @rl_dane It's a project that is stoked by management types, because they thought they would be able to replace grunt work with yes men, well there is a reason why the "grunt workers" aren't yes men, that's because of process knowledge, and their experience, now instead they have a robot that bends over backwards to their very american view of "ideas are the real work". And we see where that ends, most of the places I've seen at least has had to get people at least in the loop, or just stopping it completely, because process knowledge is important, hopefully that won't set us back too much I'm afraid it will, as they might have to build it back from scratch again.

@sotolf @paul @justin

The A.I. craze is just another way upper management demonstrates their utter sociopathy and that they are a waste of every resource they get their hands on, including human ones.

@sotolf @paul @rl_dane AI is theft and environmental terrorism.

@justin @paul @rl_dane Very much so yes.

@sotolf @justin @paul @rl_dane I do like DDG, but dislike the AI inclusion. I've started using https://start.duckduckgo.com/lite.

@cmccullough @sotolf @justin @paul

I switched to #dillo as my default web browser on all but my work machine. XD

Ain't no ai features there!! ;)

@cmccullough @sotolf @justin @paul @rl_dane

you can remove AI in the settings and save them cloud if you wish

@sirber83 @cmccullough @sotolf @paul @rl_dane they can also just remove the crap and everyone wins.

@justin @sirber83 @cmccullough @sotolf @paul

They're chasing the same shiny ball as everyone else, it seems, unfortunately.

@cmccullough @justin @paul @rl_dane @sotolf noai.duckduckgo.com

@evgandr @justin @paul @rl_dane Oof, that's even worse, in my case since it's not a tech podcast I can just jump over the most egregious parts and pretend everything is still okay, but that sounds really painful :/

@sotolf @justin @paul @rl_dane Sad but true — it is very sad

@sotolf @evgandr @justin @paul

Yeah, that's just trashed. :'(

@sirber83 @cmccullough @justin @paul @rl_dane I did that, and it resetted on me after a week or so, or every time I use private browsing.

@sotolf @sirber83 @cmccullough @justin @paul

Yeah, the best thing would be to change your browser settings to go to the noai site directly (or use a different search engine, like #clew! ;)

As for me, I'm surprisingly happy just using dillo a lot of the time now. XD

@rl_dane @cmccullough @sotolf @justin @paul erm, still those done server-side

@mirabilos @cmccullough @sotolf @justin @paul

I know, but also, the version of the page that dillo pulls down doesn't have any ai features... and dillo doesn't have any JS to enable them, anyway. ;)

@justin @sotolf @rl_dane it's also cyber-terrorism, bringing down community services through relentless webscraping

EvilCorp is everywhere.

@paul @justin @sotolf @rl_dane it’s also inherently related to TESCREAL, i.e. schools of thought that hate humanity

@mirabilos @paul @justin @sotolf

I'm not sure if #TESCREAL dopes hate humanity, or if they just deeply want to manipulate, control, conform, and terraform it to their own utterly twisted image of what it should be.

I do wish they'd all just go snog a black hole.

@cmccullough @sotolf @justin @rl_dane yeah, lite is good. Though I like thought they may be using noai[dot] to analyse how many people go to the effort to avoid AI. Lite could be argued the information is skewed by low powered devices.

Like many have said though, I wish it was the other way around, i.e no AI by default and have an ai[dot] for those that want it

@paul @cmccullough @justin @rl_dane just a bit sad that there isn't a short verison of it, as I'm so used to just writing ddg.gg having to write the full noai.duckduckgo.com is so much longer of a way, and so much more typing of uncomfortable to type words.

@sotolf @paul @cmccullough @justin @rl_dane oh c’mon

# DuckDuckGo search
ddg() {
	local _q _IFS _p=/ _a=

	_IFS=$IFS
	IFS=+
	_q="$*"
	IFS=$_IFS
	case /${BROWSER:-lynx} in
	(*/dillo*)
		# make result page and target links work
		_p=/lite/ _a='&kd=-1' ;;
	(*/links*|*/lynx*)
		# avoid automatic redirect
		_p=/lite/ ;;
	esac
	${BROWSER:-lynx} "https://noai.duckduckgo.com$_p?kp=-1&kl=wt-wt&kb=t&kh=1&kj=g2&km=l&ka=monospace&ku=1&ko=s&k1=-1&kv=1&t=debian&q=$_q$_a"
}

@mirabilos @sotolf @paul @cmccullough @justin

I used to use a script to do DDG from the terminal, but now I'm just using dillo 90% of the time. ;)

@rl_dane @cmccullough @sotolf @justin @paul search for mksh and see the summary for the linux.die.net/man/1/mksh entry. If that’s not slop I’ll eat a broom.

@mirabilos @cmccullough @sotolf @justin @paul

Probably so. They used to have summaries like that before the LLM "features," so I'm wondering where they might have come from.

@paul
LLMs have the potential to be useful. Any place where a computer needs to understand a human query and do something with that query other than generating an answer, is a place where I can see LLMs as a useful tool.

A tool that links to what it thinks as the correct answer in pre-existing documentation? Awesome (my bank does this, looping the conversation to a human if the answer was declared not helpful)
@sotolf @justin @rl_dane

A tool to interpret the spoken sentence "please give me light, I can't see" as a command to turn on the lights in the room? Yes, good idea (home assistant can do this).

Using generative AI to generate answers to questions is almost always wrong. But I do see some valid use cases...
@sotolf @justin @rl_dane @paul

@wouter @paul @rl_dane @justin There is no use case that you can't just make a better version of, you know what I do if I want light in my room, I go and flick the light switch...

@sotolf @wouter @paul @justin

I wonder if instead of slathering enormous language models on top of every little thing, applying small language models in an intelligent and targeted manner where they're actually helpful won't actually be useful.

Something like calorie counting: using a small language model to parse a sentence describing what a person ate in a day and finding the requisite parts within a nutrition database, rather than using a LLM for that and just hoping that it doesn't hallucinate something.

@wouter @paul @rl_dane @justin Also who in the hell would be stupid enough to have every conversation in their room recorded, sent out of their house to be interpreted by a proprietary LLM to be able to use an arbitrary long winded impractical command to turn on the light, if you really wanted to there are things like clap sensors, or I'm sure you can have something local that can just interpret some word for it instead of just sending everything you say in your own home to the faschy tech-bros...

@sotolf @wouter @paul @rl_dane @justin given the number of people who use alexa/google assistant/siri i think you overestimated how much people care

@xarvos @paul @wouter @rl_dane @justin Yeah, maybe people in general are less worried than what I am :p

@sotolf @xarvos @paul @wouter @justin

People who install security cameras INSIDE their house really freak me out.

"You have a camera INSIDE your house, watching your CHILDREN, and you're sending that live feed across the internet to some stupid corporation without encrypting it first? Have you been tested for lead poisoning?"

@rl_dane @wouter @paul @justin

Why use a language model when what you're trying to achieve is not a language task?

Language is a means of communicating between people.

And for calorie counting, if a person said they ate bread, should it take the calories from white bread or whole grain sunflower bread, how many slices did they eat?

Personally I think the only thing that would work well for that is basically scan the EAN-barcode on the product, have a database over the nutrient, and let the user input amount.

With an LLM you just will get random bullshit out of it, this is a task of logging precise data, not generating text, and what is an LLM? A... Large Language model.

Using an LLM for non-language tasks is like eating soup with a fork... Where did right tool for the right job go?

@sotolf @wouter @paul @justin

But parsing is a language task.

@rl_dane @wouter @paul @justin

This is a task that doesn't need parsing, apart from parsing a code128 barcode, which is optimised for computers, and computer reading.

@sotolf @wouter @paul @justin

That still requires someone to install an app, open it, and scan a barcode for every ingredient used.
That obviously provides the most accurate results, but sometimes accuracy isn't the goal.

Being able to say, "Oh yeah, I had this and this for lunch today" and just writing down a natural language query and having it well-understood and being given a useful approximate value is actually very valuable for someone wanting to start calorie counting, because most people don't stick with the exhaustive manual lookup of every single ingredient for long.

@rl_dane @wouter @paul @justin You can't just YOLO calorie measurements if you really want to achieve something with it, like reducing weight, accuracy is important.

@sotolf @wouter @paul @justin

I disagree, having an approximate value can still be very informative, and makes calorie counting much easier if the person doesn't have to stop and scan each and every item.

Anything that make the process less annoying (especially at the beginning) can be very helpful. Dieting is hard enough as it is.

@rl_dane @sotolf @wouter @paul @justin

A lot of 'uses' for LLMs boil down to "do things badly to avoid having to do them properly."

@pixx @sotolf @wouter @paul @justin

Yes, but there is a case for "do something badly where you otherwise wouldn't do it at all."