this post was submitted on 27 Sep 2024
1342 points (99.4% liked)
Technology
59419 readers
5179 users here now
This is a most excellent place for technology news and articles.
Our Rules
- Follow the lemmy.world rules.
- Only tech related content.
- Be excellent to each another!
- Mod approved content bots can post up to 10 articles per day.
- Threads asking for personal tech support may be deleted.
- Politics threads may be removed.
- No memes allowed as posts, OK to post as comments.
- Only approved bots from the list below, to ask if your bot can be added please contact us.
- Check for duplicates before posting, duplicates may be removed
Approved Bots
founded 1 year ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
What does it actually promise? AI (namely generative and LLM) is definitely overhyped in my opinion, but admittedly I'm far from an expert. Is what they're promising to deliver not actually doable?
It literally promises to generate content, but I think the implied promise is that it will replace parts of your workforce wholesale, with no drop in quality.
It's that last bit that's going to be where the drama happens
I dare my company to try it. There would be so many lawsuits in 3 years.
My company will be much better off...it's made up up of 80% value workers from India. AI can't possibly be worse than those guys at code.
It's the other way around actually https://www.businessinsider.com/amazons-just-walk-out-actually-1-000-people-in-india-2024-4?op=1
That's a prime (pun intended) example of what I'm talking about. Amazon likely hired them to write the algorithm to watch people shop, they couldn't figure it out so they decided to watch the video 24/7 instead.
That's just racism, try again
Meh...it's just a fact. You hire developers for dirt cheap, you end up with dirt cheap solutions.
They want AGI, which would match or exceed human intelligence. Current methods seem to be hitting a wall. It takes exponentially more inputs and more power to see the same level of improvement seen in past years. They've already eaten all the content they can, and they're starting to talk about using entire nuclear reactors just to power it all. Even the more modest promises, like pictures of people with the correct number of fingers, seem out of reach.
Investors are starting to notice that these promises aren't going to happen. Nvidia's stock price is probably going to be the bellwether.
You obviously don't get what ai is or what it's potential is.
And neither do you.
Lol
Accuracy, consistency, explainability.
It delivers on what it promises to do for many people who use LLMs. They can be used for coding assistance, Setting up automated customer support, tutoring, processing documents, structuring lots of complex information, a good generally accurate knowledge on many topics, acting as an editor for your writings, lots more too.
Its a rapidly advancing pioneer technology like computers were in the 90s so every 6 months to a year is a new breakthrough in over all intelligence or a new ability. Now the new llm models can process images or audio as well as text.
The problem for openAI is they have serious competitors who will absolutely show up to eat their lunch if they sink as a company. Facebook/Meta with their llama models, Mistral AI with all their models, Alibaba with Qwen. Some other good smaller competiiton too like the openhermes team. All of these big tech companies have open sourced some models so you can tinker and finetune them at home while openai remains closed sourced which is ironic for the company name.. Most of these ai companies offer their cloud access to models at very competitive pricing especially mistral.
The people who say AI is a trendy useless fad don't know what they are talking about or are upset at AI. I am a part of the local llm community and have been playing around with open models for months pushing my computers hardware to its limits. Its very cool seeing just how smart they really are, what a computer that simulates human thought processes and knows a little bit of everything can actually do to help me in daily life.
Terrence Tao superstar genius mathematician describes the newest high end model from openAI as improving from a "incompentent graduate" to a "mediocre graduate" which essentially means AI are now generally smarter than the average person in many regards.
This month several comptetor llm models released which while being much smaller in size compared to openai o-1 somehow beat or equaled that big openai model in many benchmarks.
Neural networks are here and they are only going to get better. Were in for a wild ride.
My issue is that I have no reason to think AI will be used to improve my life. All I see is a tool that will rip, rend and tear through the tenuous social fabric we're trying to collectively hold on to.
A tool is a tool. It has no say in how it's used. AI is no different than the computer software you use browse the internet or do other digital task.
When its used badly as an outlet for escapism or substitute for social connection it can lead to bad consequences for your personal life.
When it's best used is as a tool to help reason through a tough task, or as a step in a creative process. As on demand assistance to aid the disabled. Or to support the neurodivergent and emotionally traumatized to open up to as a non judgemental conversational partner. Or help a super genius rubber duck their novel ideas and work through complex thought processes. It can improve peoples lives for the better if applied to the right use cases.
Its about how you choose to interact with it in your personal life, and how society, buisnesses and your governing bodies choose to use it in their own processes. And believe me, they will find ways to use it.
I think comparing llms to computers in 90s is accurate. Right now only nerds, professionals, and industry/business/military see their potential. As the tech gets figured out, utility improves, and llm desktops start getting sold as consumer grade appliances the attitude will change maybe?
That is a miopic view. Sure a tool is a tool, if I take a gun and use it to save someone from getting mugged = good if I use it to mug someone = bad
But regardless of the circumstance of use, we can all agree that a gun's only utility is to destroy a living organism.
You know, I know, everyone here knows, AI will only be used to generate as much profit as possible in the shortest amount of time, regardless of the harm it causes. And right now, the big promise of AI is that it will replace costly human employees, that's it, that's all.
Fortunately, it is really bad and unlikely to achieve this goal
A better analogy is search engines. It’s just another tool, but
When I started as a software engineer, my detailed knowledge was most important and my best tool was the manuals. Now my most important tools are search engines and autocomplete: I can work faster with less knowledge of the syntax and my value is the higher level thought about what we need to do. If my company ever allows AI, I fully expect it to be as important a tool as a search engine.
And this is when the cost calculation comes into play. Using a search engine is basically free, using OpenAI for development is tied up with licenses and new hardware.
So the question will be, are you going to improve efficiency to the point where the cost of the license and new hardware is worth the additional efficiency?
Currently my company is more concerned with intellectual privacy, security, liability. Of course that means they’ll only allow ai where they can pay for guarantees, and that brings us back to the cost.
Does it though?
They promised no programmers needed in 5 years. (well not promised, somebody did say that but not OpenAI staff, I think). The cost of AI both in money and energy use, does not really justify the limited aid it can provide to a programmer. You are never getting enough additional efficiency from said programmer to justify those costs
Even more hated than when every customer centre moved to India
Again, at that cost? the marginal improvement does not add up
Is it though? if I can only trust it with answers I already know enough to discern whether I am getting bullshit or not, then it's not worth it. As it it today, I cannot trust it with any search I really do not know the answer to (or can easily verify) as it can be throwing complete bullshit at me and I would have no way of knowing either.
Again? you mentioned the processing docs already... but again I tell you, who will pay the heavy costs just so internal memos are written slightly better? and everything your company sends out would have to be reviewed as you do not want AI promising something you cannot deliver via hallucination
You keep mentioning cost, and in the grand scale of "there's no such thing as a free lunch" there's a large cost but for users, they're just paying for a license from Microsoft to have copilot in their visual studio software or in M365 apps, etc.
So for helping with development, it's really not that expensive for the users. Also, "they" make lots of ridiculous claims, and i don't know who said it, but no developers in 5 years is a wild claim that no one should've thought was real.
It's expensive enough my employer (of more than 2000) decided to only trial it with a small subset of seniors. It's not just the license, it comes tied up with new hardware
So far nobody likes it. Most people use it to summarize meetings and we just got a memo saying we need to review the summaries because it keeps missing important data
Having said all that, when I mentioned the cost, I was referring to the cost of training the models. And without a proper business plan to monetize it, it's is still unclear how this version of AI could be actually sold for profit.
Remember that cost, is not just a number. It's the number in relationships with the benefit it provides.
For OpenAI, it has yet to produce profit that is not just venture capital and for us as user (us, I cannot speak for everyone) it has not saved us a dime after getting expensive hardware and licenses
Oh and for the final point. True, openAI may not have been the one to say no programmers in five years although, replacing people has always been their angle. But by now we have seen OpenAI play so fast and loose with all their claims and benchmarks, we cannot believe a word they say (which you seem to do and keep on posting here).
I've only made the comment you're replying to. I'm not whoever you're thinking.
Yes, my bad, apologies
I thought you were the person I replied to originally
All good. I think we're thinking of this from different aspects anyway. I'm thinking a company just subscribes as part of their office subscription and Microsoft is doing the heavy lifting of the cost and hardware. I don't know how OpenAI makes money besides their little subscription.
As far as I have read, that's it, which is not profitable. They have been coasting on Venture Capital only so far.
Yeah, I know better than to get involved in debating someone more interested in spitting out five paragraph essays trying to deconstruct and invalidate others views one by one, than bothering to double check if they're still talking to the same person.
I believe you aren't interested in exchanging ideas and different viewpoints. You want to win an argument and validate that your view is the right one. Sorry, im not that kind of person who enjoys arguing back and forth over the internet or in general. Look elsewhere for a debate opponent to sharpen your rhetoric on.
I wish you well in life whoever you are but there is no point in us talking. We will just have to see how the future goes in the next 10 years.
Lol oh the irony