I agree, however, one could use the same argument with what we learned. Put another way, how do we know what we learned is accurate and/or true? Yes, we have to put our faith somewhere otherwise we're just spinning our wheels, but the rationale still stands. Humility in the face of being wrong and the constant pursuit of truth is the best moral baseline for growth. That, imo, should be the constitution of AI.
Artificial Intelligence: How are you using it in everyday life?
- Thread starter isucyfan
- Start date
I agree, however, one could use the same argument with what we learned. Put another way, how do we know what we learned is accurate and/or true? Yes, we have to put our faith somewhere otherwise we're just spinning our wheels, but the rationale still stands. Humility in the face of being wrong and the constant pursuit of truth is the best moral baseline for growth. That, imo, should be the constitution of AI.
As an advertising Creative Director I use AI to comp frames for TV/Digital spots to sell into the client. A lot of times I use AI in conjunction with Photoshop to get the exact look, feel, and composition I desire. It’s pretty slick and a huge time saver. I’m looking to move into more motion based AI as well. Ex. Firefly
For writing scripts, AI isn’t quite there since is can’t pick up on the nuances of being human, like humor or emotion, but it is good for proofing content and coming up with something that might spark an actual good idea. That’s still pretty rare though.
Later today, I’ll be using Midjourney to generate a specific illustration style for a logo design. I’m curious to see how close I can get it to actually producing what I’ve got for reference/composition. Regardless, I bet there will be enough there to save me a fair amount of time during the process.
AI is amazing, but not the end all be all that it will eventually become. We have a crazy, and unpredictable, future ahead of us. I can’t wait to see how it all unfolds.
For writing scripts, AI isn’t quite there since is can’t pick up on the nuances of being human, like humor or emotion, but it is good for proofing content and coming up with something that might spark an actual good idea. That’s still pretty rare though.
Later today, I’ll be using Midjourney to generate a specific illustration style for a logo design. I’m curious to see how close I can get it to actually producing what I’ve got for reference/composition. Regardless, I bet there will be enough there to save me a fair amount of time during the process.
AI is amazing, but not the end all be all that it will eventually become. We have a crazy, and unpredictable, future ahead of us. I can’t wait to see how it all unfolds.
I used ChatGPT yesterday to assist in putting together a workout program for me. Did the first one this morning. Pretty freaking impressive IMO
I wrote an article on this for work. You definitely want to remove any intellectual or proprietary information, any PHI, anything of the sort. It is important to remember that it pretrains off of the information put in. As such, anonymize or deidentify your prompts and uploads to avoid giving away trade secrets.
Some best use additional items are to:
- Opt out of them collecting pretraining info from your input by turning off data sharing - even then, I don’t trust it
- Either don’t log in, or use a ghost email to avoid personal info being collected
- Always fact check anything generated
- Remember it is checking the whole internet for retraining, and keep in mind the amount of biased or just wrong info there is out there. Garbage in, garbage out. Mitigate bias and misinformation by checking output with groups who may be affected by this output.
Same. I tell Siri and Alexa please and thank you every time I talk to them, and I tell my kids to do the same!
Last edited:
This is just my opinion but... while all of this is true and good practice... humans are lazy AF. AI makes things easy. There's no world in which a written policy of good practices (as you listed) would actually prevent from PI/IP/PHI data exposure.
I'd prefer the ability to run modals locally or within a company's network. There's obvious drawbacks of that, but it would seem to me to be a better security practice.
I absolutely agree - not everybody is going to follow best practices. Which sucks. It’s how we have current PHI leaks with address issues or people misstuffing envelopes. When people take shortcuts, we all suffer.
I just read every post. And, as someone doing pseudo AI since the 80's, I can see why 'fear' is prevalent. Heck, here at a late age, I even fear it. Most people misinterpret great computer tools for AI, but real AI learns from its self accumulated dataset. For the sciences, this is an incredible tool. For trades like marketing, law, business, programming, politics, etc, I see rapid growth to first supplement and then replace these trades at a rate faster than the ability to retool those folks. As several experts have pointed out, the blue collar trades will be the safest (in the near term) as the benefits of AI won't be perceived until robotics is perfected ... which is the next phase. Of course the economical fallout of trades being replaced will be scary astronomical. Me? I'm doing everything in my power to reduce my dependence on technology. ... Learning how to make a candle, or growing my own food, or canning, can be rather rewarding. I already live in the "Quiet Zone" so I'm already experiencing the peace of no cell phone. And, having also worked in that industry, I can now say I don't miss it at all.
IF there is a Matrix, we are in it now.
And what better way to fool us than to make a movie about the Matrix.
And then 20 years later make a sequel Matrix movie that is about a Matrix video game being created within the Matrix that is about the Matrix.
PS. If we ever create AI robots, just please do not make them physically really f***ing strong! Super intelligent robots do not need to be able to scale a skyscraper in seconds, sprint 100mph, lift a car, or punch a hole through a person's sternum. Give them the strength of a standard healthy grandmother. That's it.
Midjourney can create some amazing stuff, but honestly, after using it for a couple weeks trying to perfect some logo and illustration ideas, I walked away feeling frustrated and used Canva instead. Getting the prompts just right definitely takes some learning, but I constantly felt like the AI would go off on a strange tangent rather than fine tuning what was there.
I'm intrigued by this. How does this work? Say you've written some code and discovered a bug and maybe narrowed it down to one module or one function. What are the steps you take from that point to get AI assistance?
It’s pretty good, at least for data analysis which is what I use it for. I’m not an app developer so I can’t speak to that.
You can ask a question and it will give you a good starting point. I would say you need to know some general concepts of how the data is structured and you have to know specifically what you envision as the final outcome. But it can give you 75% of what you need while you fine tune the remaining 25%.
The other thing it’s good at is debugging. You can ask “what’s wrong with this line of code and 90%+ of the time it will identify the error.
I'd be careful to double check what it spits out. I've had it conflate two different specs found in the same source.
I use it in my day to day to code. Playing with GitHub Copilot Agent lately. It's good at the rote, but will often produce things that do not work out and will give up when I try to correct it. So generally I'll have it work on some boilerplate and iterate on that till it gets stuck and then I'll take back the wheel and do the interesting things.
Management wants AI to be integrated into all our products. We'll see how it works out.
Last edited:
Better than a google search, but I'm not allowed to use outsourced AI for work. Can only use works AI and not supposed to use it to write code, just for tips on how to do things within the framework and such. From what I've seen, powerful enough that I could submit a bunch of source code and have it generate new code for me. Awesome but scary.
Want to take a stab at defining sentience in a way at that precludes inorganic implementations?
I remember watching Contagion in 2011. If you wanted a society to act a certain way during a pandemic, what better way than a star studded movie essentially mimicking the actual one? Hmmmm. Fun to contemplate.IF there is a Matrix, we are in it now.
And what better way to fool us than to make a movie about the Matrix.
And then 20 years later make a sequel Matrix movie that is about a Matrix video game being created within the Matrix that is about the Matrix.
PS. If we ever create AI robots, just please do not make them physically really f***ing strong! Super intelligent robots do not need to be able to scale a skyscraper in seconds, sprint 100mph, lift a car, or punch a hole through a person's sternum. Give them the strength of a standard healthy grandmother. That's it.
I found a new use for it yesterday. I've started using it as a GenZ slang translator to annoy my kids when I sent them text messages.
Latest posts
-
-
-
-
-
Chicago Bears bought Arlington Heights track only to build in Indiana?- Latest: FinalFourCy