A.I. is Exploding! Can We Really Trust It?
HTML-код
- Опубликовано: 29 июн 2024
- To try everything Brilliant has to offer-free-for a full 30 days, visit brilliant.org/TravisMedia . You’ll also get 20% off an annual premium subscription.
There is a mindblowing amount of money flowing into AI startups these days. Companies are taking on billions in debt to build the next AI innovation.
Yet there remain a few barriers which will continue to slow adoption.
In this video, we'll discuss those barriers and I'll also show you two simple and more matured tools that I use on a regular basis. Just because there's a lot of hype, doesn't mean the hype helps us out a lot currently. In fact, with the failings we're seeing in Gemini, it seems quite the opposite. I'll also recommend a couple of my favorite AI newsletters.
Timestamps:
00:00 Intro
00:07 Insane fundraising
00:43 AI setbacks and the future
03:56 Sponsor
05:05 Choosing AI tools
06:01 A simple AI upgrade
07:37 ChatGPT as a Jupyter notebook?
12:12 Simple is often better
Join the Travis Media community - imposterdevs.com
Udemy Deals (regularly updated) - travis.media/udemy
** Career Path Coding Tracks **
Web Developer - geni.us/jBigBd
Software Engineer - geni.us/AbMxjrX
Machine Learning - geni.us/GporLlT
Python Developer - geni.us/tv2FJBU
DevOps Engineer - geni.us/MgHtJ
** My Coding Blueprints **
Learn to Code Web Developer Blueprint - geni.us/HoswN2
AWS/Python Blueprint - geni.us/yGlFaRe - FREE
Both FREE in the Travis Media Community - imposterdevs.com
My microphone - amzn.to/3sAwyrH
** I write regularly **
travis.media
** FREE EBOOKS **
📘 travis.media/ebooks
LET'S CONNECT!
📰 LinkedIn ➔ / travisdotmedia
🐦 Twitter ➔ / travisdotmedia
🙋🏼♂️ Website ➔ travis.media
#aiprogramming #selftaughtdeveloper #chatgpt #gemini
** Some of the links in this description may be affiliate links that I may get a little cut of. Thank you. 
Наука
To try everything Brilliant has to offer-free-for a full 30 days, visit brilliant.org/TravisMedia . You’ll also get 20% off an annual premium subscription.
My business uses AI 24/7, giving me a considerable competitive advantage because none of my competitors have figured it out yet. I don't think they even know about it. However, as you mentioned, I have to actively supervise what it generates to correct errors. When I first used GPT-3, I thought all the predictions about AI's future would come true. After using it for two years, I see it's just an amazing statistical model at predicting the next word, but I'm doubtful it will ever become smarter than a human without an insane amount of computing power or some major breakthrough.
How do you use it?
@@rejectionistmanifesto8836 AI monitors all of my office messages, emails, and phone calls. It forwards all of the urgent client messages to the support team eliminating wait times. All meetings and conversations are recorded and summarized helping us have customized profiles on all of our clients. We don't let AI directly interact with clients yet due to the hallucinations.
@@rejectionistmanifesto8836competitor detected 😂
@rejectionistmanifesto8836 chat bots I guess
@@johnyepthomi892 lol
*_AI is not really here yet. All we've arrived at are smart predictive LLMs. It may be years before we reach artificial intilligence, let alone an AI super intelligence_*
This is such a good rundown. The whole AI rollercoaster has been a wild ride so far
You ain’t seen nothing yet
Agree. Using AI for coding does help with syntax but it's annoying because it keeps giving incorrect suggestions.
"Modern MBA" has a really insightful video on the topic
Great breakdown. Thanks, Travis!
The idea that AI will create new jobs is a bit of a contradiction if the point of AI is to duplicate or exceed the cognitive abilites of the average person- which is the declared aim of every large AI developer- they want 'AGI'- with the 'g' here standing for 'general'. So artificial general intelligence.
This may never actually happen for all sorts of reasons but if it does then-by definition- any new job created by AI will itself be done by AI, because at that point AI will match or exceed the abilites of the average person.( Which is- of cours- most of us!)
So while it's true that in the past new technologies have given rise to new and unforseen jobs this may not hold true in the case of a technology that is explicitly being developed to replace not just expertise but the cognitive capacity from which that expertise arises. There is-after all- only one purpose for creating a general as opposed to specialised Artificial Intelligence and that is to replace the average person in the economy with a cheaper alternative.
most people have yet to realise that when Sam Altman says he is going to develop AGI what he means is that he wants to create something as functionally smart as the average person in order to put them out of a job. Mass unemployment is not a side effect of Altman's project- it IS Altman's project- and the same is true of all the other AI developers. When they say they want to help mankind they don't mean ordinary people living today- they mean some abstract future people- not you and me- us they want to screw over for profit.
The joke is that should they succeed the outcome must surely be a collapse in demand as people lose their jobs to AI- so where is all the profit going to come from to pay back all those investors?
The right question is, where they want to apply AI, food production? Energy production? They want me unemployed so I cannot eat and give my land to nature after I die? These are the right questions. If it's just for archiving and query information (the state that AI is mostly known today) then I don't see any problem with it. If this tech is for screwing people and installing some type of dystopian future, I am not the one that is going to be screwed, they will be the ones that are going to be screwed. There is no escape for them, they will have to face me and others at some point in time and it will be really painful on their side.
Its ok ubi will be here either at end of this year or next year
Another way to look at it is a post labor economy we could go to a resource based economy. So be free to do your passions and interests or use ai to make money such as anything you create in business or entertainment.
Reason is this is the last Industrial Revolution on Earth the machines are catching up to us.
Yes something like that. But not all jobs will go away just enough to cause a problem to mass unemployment so it depends what you do.
with today's sponsor, Beeliant
I agree that these AI tools can be useful and I appreciate your well crafted demo of ChatGPT 4.0. I have to say though that even basic Excel/Numbers etc. skills would accomplish all of these actions much faster than having to type in sentence structure queries with the exception of establishing trends. Honestly though why would you want someone else telling you which trends are important to you? There is value in thinking about the data matters to you.
Not Trust but responsibility and liability
The next employer will ask Travis: how do you sort this data? Travis will answer: Let me to ask ChatGPT.
Definitely can’t do that in an interview
I trust AI more than my wife 😂. And its nice to talk to
😲
Thanks for those great insights sir 🙏 👍 🙌
Irony is the google holding largest amount of data of the world and its ai model is ridiculous
Gemini 1.5 Pro and Flash are Great though !
I think we can trust AI but at the cost of making folk lazy which isn't a good thing.
We cant trust it as the humans programming and controlling it have shown they are like all other humans where power corrupts them.
❤❤❤❤❤❤❤❤
People miss the main point every time!
Artificial means - FAKE, DUPLICATE, COPY...
YOU WOULDN'T eat Artificial food all the time.
The only reason people are still talking about
This parlor trick is because they think they
Can use this to make some easy money!
"Artificial" means human-made, not fake. LLMs are designed to generate text based on real data.
Unlike synthetic food, LLMs don't impact physical health. Many artificial products, like medicines, improve life, just as LLMs enhance productivity and creativity.
LLMs are sophisticated and have practical uses in language translation, content creation, customer service, education, and more.
LLMs provide genuine benefits beyond profit, democratizing information access, automating tasks, and improving business interactions.
The AI community is actively working on ethical guidelines to mitigate risks like bias and misinformation, ensuring responsible use of LLMs.
@robosergTV But? The real explanation still means NOTHING.
People live and interact in a REAL world..
Machines live and act thru computers...
It's a virtual world...an imitation of the real thing.
But what if the Imitation is as good as Real ?
AI Generated Text by a Frontier Model with Good Prompting is indistinguishable from Human Written Text.
One can even give examples of a specific writing style to generate a new one on that writing style.
The texts which look obviously AI Generated are either from an Old Model, or didn't have good prompts, or usually a combination of both.
Even the AI Text Detectors are almost useless.
There's no way one can detect a good AI Generated Text from Human Written Ones.
@srikrishna2561 Our fathers and grandfather's got along fine without artificial intelligence and so can we.
When WE accept anything from big tech as a standard, WE give them the authority to change whatever they want - whenever they want.
Can't follow your Argument. Our brains are inside our Skulls, just connected by a few sensors to the outside world (eyes, ears, skin) with some moving parts.
It is fairly easy to make a robot with the same sensors and motion to connected a digital brain to the physical world.
Given, we are not there yet. But there is no law of physics preventing this from happening.
Glue on pizza? That's actually bettter responses than I typically get. The ai:
1: Makes up methods.
2: Mutilates code.
3: Solves problems by simply creating other problems.
4: Can't understand basic context.
I have stopped giving it even simple problems like "split this string such that 80 line length is observed" and it can't do it. It takes 3-5 attempts on the newest GPT....
It got WORSE. At least I could have it do boilerplate. Now the boilerplate is sussy too....
Wait for devin
@@ArtificiallyClever that is also shit. Maybe it will never see the light.
Wait till the end of the year how much more advancements coming and job losses to ai
"but perhaps we're not asking the right questions" this is poppycock. It has non deterministic behaviour.