Jeffrey Goldberg
Al Sweigart"The dumbest person you know is being told 'you are absolutely right' by an AI right now."
SnoopJ
Ozzelot 
@SnoopJ
I'm gonna have to be very very healthy.
(Judging from my family, we can hang on into good age with minimal intervention, so hopefully that's gonna happen.)
@AlSweigart
I'm gonna have to be very very healthy.
(Judging from my family, we can hang on into good age with minimal intervention, so hopefully that's gonna happen.)
@AlSweigart
pmjv s klapkami na očích@pmjv
(many memes say the secret to long life is spite, alcohol or avoiding people. i'm healthily dosing at least one of those at any given time.)
@SnoopJ @AlSweigart
(many memes say the secret to long life is spite, alcohol or avoiding people. i'm healthily dosing at least one of those at any given time.)
@SnoopJ @AlSweigart
so that's the catholic trinity i've been hearing so much about
CC: @[email protected] @[email protected]
CC: @[email protected] @[email protected]
@pmjv
cat-holic... yep, liking cats may help, or it may not, but it's nice to meowmeow.
@SnoopJ @AlSweigart
cat-holic... yep, liking cats may help, or it may not, but it's nice to meowmeow.
@SnoopJ @AlSweigart
pastryman
Okuna@ozzelot @SnoopJ @AlSweigart once a doctor told me I have to give up drinking, cigarettes and too much meat.
I never drank, never smoked and was vegetarian.
Now I am prepared for the next doctors visit to oblige to his request
:-) SCNR
Lizzie@AlSweigart I am not!
Critical Angle@AlSweigart I dont know where this idea comes from, because my AI (Lumo) corrects me all the time.
People should just keep their queries fact-based and don't ask it if Anney or Chrissy should be should be their best friend. To be honest I'd just agree with whatever they said, as well.
Elizabeth Sudduth@criticalangle @AlSweigart chatbots don't know what "facts" are
europlus 
@AlSweigart @jwz I’d be in so much trouble if I shared that in some of my circles…
…am still considering it.
Demokritus Jorik 
@AlSweigart I almost thought you are an AI because your display name.
Wulfy—Speaker to the machines
kasperdOne needs to be able to judge whether the output from an AI is trustworthy or not. Anyone who cannot correctly make that judgement isn’t smart enough to be using AI.
@kasperd @n_dimension @AlSweigart in my AI classes I usually start with this:
I dont need AI.
If I have no clue, I can’t use AI because I am not sure it is lying. If I know the topic, I dont need AI anyway.
:-)
Sometimes I know enough about a topic to verify the correctness of an answer. In those cases there is less than 10% chance an AI can give me a useful answer to the question. But there is a non-zero chance.
One example from my own experience was that I wanted to know the git equivalent of this command hg log -r 'children(.)'. I have searched for an answer using search engines. I have asked some people who are promoting use off git. All of that lead me to suggestions that would only produce incomplete outputs. And for years nobody I asked could suggest a fix for this.
Eventually I asked an AI, which initially gave me the same answers I had seen before which did not work. After pointing out 2-3 times that the suggested solution didn’t work, the AI eventually managed to give me a useful suggestion. It hadn’t produced a fully working solution, but it had given me a different approach that I was able to turn into something working by myself. The key point was to use the command git cat-file --batch-all-objects --batch-check as part of my solution.
I wouldn’t have managed to find that solution without the help of AI. But using the AI certainly wasn’t a quick and easy way to a solution either. I had to go through loads of useless output from the AI to pinpoint that one tiny piece of information that was useful.
So my conclusion is that current LLM are sometimes better than nothing, which is a pretty low bar to meet. And they are very far from the all-knowing infallible oracles that proponents say they are.
The free models are dumber.
Their sole purpose is to get you hooked and enter the lunapark where you pay the full price for the mechanical turk...
...whenever you hear an #Ai noob say "Ai is sh*t because it gives rong answers", chances are they are using, what I call the "sideshow" model.
In 95% of the cases when a prompt is given, and the person complain how dumb the model is, I have found my commercial, fronteer model, does not make that mistake.
@n_dimension @AlSweigart @kasperd
I mostly use local AI. The new models are really good and for my needs sufficient.
I have never paid for Claude, chatgpt or Gemini or used their apps or used them on their own websites.
If I have the need to use a bigger model (which gets less and less) I use Venice.ai. I would donate to them anyhow because they have privacy in mind. And they offer a wide variety of models to choose from.
When people tell me how dumb AI is, I asked them how they engineered the prompt. In 98% of the cases they have no clue what I mean because they use the models as Google Search.
I mostly use local AI. The new models are really good and for my needs sufficient.
I have never paid for Claude, chatgpt or Gemini or used their apps or used them on their own websites.
If I have the need to use a bigger model (which gets less and less) I use Venice.ai. I would donate to them anyhow because they have privacy in mind. And they offer a wide variety of models to choose from.
When people tell me how dumb AI is, I asked them how they engineered the prompt. In 98% of the cases they have no clue what I mean because they use the models as Google Search.
bgoodwin@AlSweigart Corollary: The most evil person you know is being told “You are absolutely right!” By an AI right now.
geolaw
mHtt@AlSweigart its like using a cheese grater on your brain
internet george@AlSweigart 1-bit LLM “A”: exactlyyy
1-bit LLM “B”: exactlyyyyy
1-bit LLM “B”: exactlyyyyy
aibyweb. @AlSweigart 😌
Probably because, they only seek what they want for the time being. If anyone use this thought process to their advantage, the very same Ai becomes much powerful, rendering outstanding responses for us.
�@AlSweigart "should we raise taxes"
w7comHey! We all need validation.
O.Voãoyou are absolutely right
sean_devs@AlSweigart you're absolutely right
CressCompare the rather obvious "baiting" by AI's with the convoluted lies and manipulations by humans at the moment on the world stage.
I'd much rather have an AI running the show as long as its trained only by compassionate and aware human data, themselves elected from society.
Then have the civil servants carry out all the work and offer in AI guardrails. Keep the ego's and lies out of leadership! Bring back Transparency and Integrity.
Just an Idea 😉
slotos@CressSalad @AlSweigart Is this the new version of „blockchain elections”?
Did not know about this until I read an article - thats interesting:
https://www.blockchain-council.org/blockchain/blockchain-based-voting-systems/
Personally I think we just as a species/planet need to get human emotions out of the picture in making big decisions - like digging up all the last oil, starting wars etc. We have to have some level headed advice and decisions, OK, we can put human gate-post on it. But it should never be 1 individual or party. I would even say for local issues it can be left alone!
Mu@CressSalad @AlSweigart AI can't run the show it can just return statistically generated answers.
Isnt that what we want (zero ego, conflict of interest, divide and rule, misappropriation of funds, war mongering errr 🙂 anyway its an idea, seems way better to have a neural net making choices than certain people at large in the world at the moment)
@CressSalad @AlSweigart LLMs aren't neural nets.
In addition, organisations come into new and novel situations all the time, LLMs can only go on their training data. An LLM will not be able to provide any useful information in that scenario.
Leading an organisation is a skill-based job. In the LLM scenario, the CEO would be the person doing the prompting.
Further, if you are setting up a scenario where everyone in the organisation does what a LLM tells them to, you're very quickly going to have a real problem.
mathew@CressSalad @mu @AlSweigart A big problem with statistically generated answers is that correct information isn’t necessarily the most statistically prevalent. In fact, there are very well funded organizations dedicated to making sure that it isn’t. Yet people perceive information from computers as being both unbiased and more accurate than information from humans.
True, this helped me understand:
https://www.ibm.com/think/topics/ai-vs-machine-learning-vs-deep-learning-vs-neural-networks
I am thinking though in a few years we should be able to ask questions of these systems, what ever level of AI we get to, to example draw up a national budget or list natitonal priorities for the population (for example health care vs military), not just to the benefit of some prime minister or president. Maybe we are already running these models. I dont know.
Yeah like I said it needs to be thought through its not a comment about a micro-architecture, need to use some creativity - imagine taking learning from the best minds in society, could be for example, gov has 10000 objectives, then the people in society are polled and paid for their advice (lets say 500k people). Then solutions that come back are used in training. Objectives will be balanced by data from the society and environment using scientific bodies and papers
My orginal comment was just to take out the fckn ego crap from leadership - surely "smart" people can work out how to do that using modern tech, and still have a lovely country to live in and not have so much waste.
I am not saying I have all the answers, only I see a major issue at the moment - thats pretty obbvious right? LOL
@CressSalad @mathew @AlSweigart almost everyone I have ever met that has worked for a government department has genuinely wanted to help people. A good 50% of politicians I have met have also genuinely wanted to help people.
There are bad eggs everywhere, but you seem to be saying that all government decision making is bad, and that is incorrect.
As rough as the government can be, private sector is worse. What you want is an assemblage of people who have the time and expertise to understand the problems affecting us, and working to solutions.
That's what the government is.
Individuals, usually leaders, can be horrible and make bad, destructive, decisions, but that's a problem with voters asking those people to make those bad decisions, it's a feature, not a bug.
By the time you have an expert in the issues who can program something to make good decisions, you're going to already be on the next problem.
I get the desire, the yearning, to have a better system than we have, but you appear to be trying to solve a people problem with technology, and that's never going to work.
Yeah you know Mu, I think your from NZ right - you are blessed with good govourning there and dont seem to have a serious right wing issue
I lost count of bad eggs in Europe and the UK and now the US in the last decade
So maybe the real problem is education - school at home or society in general
I dont want to get rid of people and totally get people go into politics from a perspevctive of care. But politics in the North seems art of division not togetherness IMHO
@CressSalad @mathew @AlSweigart we have right-ring problems here, but we look to places like France where people will burn the city down if the government does stuff they don't like, the European data laws and the Scandinavian social nets as inspiration.
Frankly, there is a lot to emulate up north, and we have anti-vaxxers and anti-trans people in government right now. We aren't a haven we're just getting to the same place a bit slower.
Codey McCodeface@mathew @CressSalad @AlSweigart @mu
“On two occasions I have been asked, 'Pray, Mr. Babbage, if you put into the machine wrong figures, will the right answers come out?' I am not able rightly to apprehend the kind of confusion of ideas that could provoke such a question.”
bbbhltz@AlSweigart A person I know quit their job last week because AI told them the contract was illegal. It wasn't a great contract, but it was legal.
@Okuna @AlSweigart My students told me about it and my mother in law complained about it to my partner. There was certainly a moment of panic. At the time my mother in law was using ChatGPT to prepare a legal thing so she was upset.
The Turtle@AlSweigart it used to be you had to get drunk in a bar to have that happen.
Stoneface Vimes@the_turtle @AlSweigart that was much more enjoyable though.
Harald K.@AlSweigart It's democratization. Powerful people always had access to people who told them "you are absolutely right".
Andrea@AlSweigart if that's not scary I don't know what could be.
Robert Kist 🇦🇹🇸🇬@AlSweigart fake news. Trump is too dumb to use an AI