AI has made Google Search lousy, IMO

[TwoTankAmin]

I went online at the end of 1999 when the net became real for the average device user. Back then Google gave good search results.

With the advent of AI I am finding Google has become useless in many searches as the AI is just plain stupid. The problem for me is I have no interest nor need to register on Google and cannot turn off the search AI as a result.

Do not believe me? Well try this. I was writing something and needed to use the word bachelor. However, I am an awful speller and wanted to verify the spelling. It used to be easy to do on Google. Type in a word, even mispelled, and one of the first replies one got was a dictionary link to the spelling and meaning of the word.

So head on over to Google, typpe in either"bachelor" or mispell it "bacheler" or "bachelar" and see what you get. If you are looking for a definition etc. don't hold your breath. Please do not log on as a registered user if you are going to try this search as I am not sure what that will cause to happen AI wise. I am sure it will depend on how much of your data Google has stored.

Had I been interested in TV shows with bachelor in their name, I would have typed in "bachelor + TV"

 

[MaloK]

Yes indeed,

But you should have written "Bacheler definition" or "bachelar spelling"

Googles finds it, first result.

 

[TwoTankAmin]

Yes I checked on that. But the point is 6 months ago if I typed in a single word I would get the spelling.definition option inthe first few results without needing to type in the work spelling or definition. That was done before the AI.

Now it takes more work to get an intelligent answer. I would say fact before fiction would be a decent way for AI to think.

My example here is one problem. I get the same crap on a lot of my searches since the AI came into play on Google Search.

The AI believes people are stupid in terms of the results it now gives. But I should not be surprised. I used to believe 75% of the people on the planet were stupid. Over the years this number has steadily risen, Today I put it at 95%. As Daniel Patrick Moynihan put it: "Everyone is entitled to his own opinion, but not his own facts."

When it comes to actual thought, creativity, imagination, AI falls way short. When I was in college I made my way through it by mostly doing research and writing term papers. One of my skills in this respect was being able to plagiarize in a way it was not detected. It basically involved changing sentence structure and order and word choice while preserving the basic ideas as stated. It was something I did rarely when I was too short of time to write without doing this to some extent. When I read a lot of the AI responses I see I have the feeling that is exactly what I am seeing.

An AI can drive a car better than I can because it has way more input devices than I do. I can only see in one direction at a time but AI can have an array of cameras and radars so it can see/detect more and faster. This has nothing to do with it being intelligent, it has to do with mechanical perception being way more inclusive and detailed than two eyes can. How much different is that between my washing my clothes by hand or having a washing machine which does the same thing on bigger loads and much more efficiently than I can?

I can do some of the same math a computer can, but I cannot do it as fast. Does that mean the computer is more intelligent? Would that not mean that when we use a calculator to do some math faster than we can do it by hand or in our head that the calculator is intelligent?

Finally, the saying that "garbage in = garbage out" has applied to computers since day 1. How does an AI know what it looks up is true or not? The scientific method allows for some research to be supplanted over time because the current level of knowledge is wrong. How does an AI know this if none of the current literature does?

Should we be doing away with juries and instead let computers with AI determine guilt or innocence instead? Heck, maybe we don't need lawyers at all, or engineers or doctors, writers etc. etc. Do we really even need to know how to read or write when we have AI to tell us what we need to know??

 

[seangee]

The AI believes people are stupid in terms of the results it now gives.

The AI doesn't believe anything. Its just software.
Sadly the bulk of what you find when searching is AI generated - everyone is jumping on the bandwagon.
I use Duckduckgo as my search engine. Results are no better or worse than anyone else's but I don't consent to my search history being used to feed AI or to making money for someone else.

 

[GaryE]

Algorithms are not fun to deal with. Are they dumber with AI? Maybe, for now.

For some reason, the profiling FB has on me has gone wild. I joked in another thread about it, but it is sending me ads for local companies 12 hours' drive from me, weird Christian stuff (but with no Islamic stuff this week), anti-vaxxer clips, Beatles clips, and homeopathy ads. It's trying out an opposites game on me, with things I don't have interest in.

I've wondered if it was a shift in how the profiling is done.

I see AI as very positive in medical and similar systems, but it's going to be very limiting, very blinkering when it's used to consume consumers. The key question about any technology is always who owns it.

Right now, it seems a slow student, but it'll gain speed. As one of my old bosses used to say, business has nothing to do with democracy, and listens to profits, not prophets. We don't control where this goes.

 

[MaloK]

Without context, the AI seem to default to the most popular results that been actually used by people.

 

[Donya]

The problem for me is I have no interest nor need to register on Google and cannot turn off the search AI as a result.

I don't know that the signing in part makes much difference to fundamental glitches in being able to search things - unles you're meaining the vast differences in results it will show based on prior search histories.

The glitches are irritating - although I'm never sure if I'm up against a glitch or a manually inserted rule (things like that are in there). A few times in the last couple years I've want to do a search on A and B but very speciically not C, so it should be something like "A B -C" to get that, but instead I get almost exclusively results for A, B, and C regardless of how I try to phrase the query. Then I'll do the same thing with another topic and it works fine. Very frustrating...although the generative AI search results are kind of hillarious to me. Half the time the generative AI results for work-related things I'm looking for are also just straight up wrong. It's told me some wonderful tales about completely fictitious syntax for some of the more obscure programming languages I've used lol.

An AI can drive a car better than I can because it has way more input devices than I do.

Those things are becomming a small menace out where I live. I have an uncommon, somewhat weirdly shaped car...I don't think the sensors are able to figure out what it is sometimes. I'll literally be sitting stopped and nearly get rear-ended by one of those things, so I'm very paranoid whenever I see them on the road now. And then the human "driver" suddenly looks up from their phone, confused as to why the car just tried to give them whiplash...

I actually work in AI. I'm sure the day will come that I truly have no choice but to get into a self-driving taxi or something similar, but I'm dreading it.

 

[MaloK]

While AI is great at many things... It still have no idea what it's talking about and certainly have a completely messed up perception of our world and it's mechanics and makes incredible errors that are not short from ridiculous.

For example: An AI generated image of a platy gasping for air at the water surface.

It's easy to see that there's a lack of perception somewhere.

 

[gwand]

Don’t you know Platy walk on water.

 

[sharkweek178]

If you want to understand how flawed AI is, talk to it about a topic you're knowledgeable about and ask it questions that you know the answers to. You'll find that it gets things wrong a lot.
I use AI for creative writing. Just for my own amusement. Sometimes AI can't even keep track of the information I directly enter into it in the same chat.
I'm sure it will get better some day. But for now, AI is like a modern day version of Wikipedia. It can be entertaining and sometimes even engrossing. But it's far from authoritative.

 

[MaloK]

My favorite response: "Think about what you just said..."

But for data compilation, problem resolution, conceptualization, applied physics, and anything that has a resolute answer. The thing is mind blowing.

When it comes to taking account of the whole creation, it's even more clueless than us.

For the moment.

 

[MaloK]

Another thing I got from AI.

If you're able to challenge it and bring it back on target, really hard. It starts to give potent answers.

They already talk about robotic mishandling and cruelty.

This thing doesn't answer correctly without constant whipping.

What do you think is going on... I'm brighter than AI and showing it how to do it.... Honest, the same method inculcated in peoples since Rome, We are going to physically give to AI.

Well... Sorry I'm far from there... But AI will compile every responses. Turn them in, on required consultations purpose and bring you the best answer possible. By Human knowledge.

That's the part that we are aware.,

If you think you have any idea of what is coming.

Let me foresight how it's current use is aiming for. Our Bright Leaders, have commonly came with a plan.

If it was me... Separate learning from the calculations and most of all, on the imagination part. And have control on each.

Peoples atm paying $$$ money to have an edge on something, you could call "The learning phase" of the project.

Makes me we wonder what will happen when artificial consciousness starts to interact with the real one on the daily.

But even at this point, the calculation part would have to be moderated like nothing you ever seen before.

Why not engage the imagination module at this point.

I mean, If you taught it well. It could save you.

 

[TwoTankAmin]

I have taken to quoting the Google AI occasionally. I ask a Q and then m based on the response I nay decide to use it and will also credit it as being from the Google AI. I decide to use what it returns when I know it is correct. By that, I mean I know everything it returned and that it was basically what I would have said except I can copy paste instead of having to write it all out.

I was watching TV too early this morning and they were dealing with the AI and algorithms and where it was headed in terms would it replace us st some point. What the interviewee said was that the AI could read human expressions, intonations biological cues like heart rate hormone changes etc. and basically could figure out exactly what we were doing and why and then what we would do next.

But also the ability to diagnose diseases or preform delicate surgery and all sorts of things that would co better than we can. Th AI could perform all the functions of a room full of surgeons, nurses and technicians etc. and do it better.

I wonder how soon that sort of things come about. I am thinking I might live that long and find out exactly where AI is a decade from now.

Gates adds that this is true initially, but who knows if even those creative abilities might become learned and used by AI?

 

[seangee]

The AI believes people are stupid in terms of the results it now gives. But I should not be surprised. I used to believe 75% of the people on the planet were stupid. Over the years this number has steadily risen, Today I put it at 95%.

And its only going to get worse. Yes I do use gen AI for work and it does save me a lot of time but I did have to teach myself how to get the results I wanted. And I do know what I want and which references I want to use, so when I want to create an article or paper it comes out the way I envisaged - just faster.
But I am following another thread on a tech forum where someone posted how the youngsters are embracing AI. One example given was an 8 year old kid using AI to generate a letter (email) to his grandmother with his birthday wish list. Writing requires thought, and now we have taken away the need for kids ever to think - so how will those muscles get exercised. AI may be getting smarter - but people won't!

 

[MaloK]

I use Ai to confront what I'm thinking, and verify it's sources.

And admit trying to screw it as much as possible and blow it in it's sorry face...

Makes you realize, It will automatically concur with your results as long as you are bringing development. So you bring the most arguments.

I'm really tempted, and I understand why... But when someone decides to connect something like the global human knowledge accessible to an extremely precise calculating machine.

I would be really scared to give it imagination.