The current iteration of BingAI really won't push it as much as ChatGPT will. I used the same prompt again, and it maybe made it 3-5% wordier. I asked it to rewrite it and "really ham it up" and it produced basically the same text again. MSFT seems to have tighter reins on it's AI output.
The top 20 jobs that GPT-4 will replace, as written by GPT-4
Software engineer not on the list.
Just like with Google and Stack Overflow, the difficulty isn't in copying and pasting the output, it's knowing what output to copy, and where to paste it.
I would agree with a few of those, but some, probably not.
It could be very useful in the customer service rep job, but mostly as a data gatherer, and possibly to cover some of the preliminary Troubleshooting steps, before handing off to a live body.
__________________
“Logic is a defined process for going wrong with Confidence and certainty.” —CF Kettering
I would agree with a few of those, but some, probably not.
It could be very useful in the customer service rep job, but mostly as a data gatherer, and possibly to cover some of the preliminary Troubleshooting steps, before handing off to a live body.
One of the dreams in a recent job was to use existing solved customer tickets to generate automated responses for similar cases, since most customer issues were variations on only a few problems, but they required a lot of effort to debug.
We didn't have the tooling in place, and only with the latest rounds of GPT would I even think we'd have a chance to get it to work well enough to show to customers.
I would agree with a few of those, but some, probably not.
It could be very useful in the customer service rep job, but mostly as a data gatherer, and possibly to cover some of the preliminary Troubleshooting steps, before handing off to a live body.
One of the dreams in a recent job was to use existing solved customer tickets to generate automated responses for similar cases, since most customer issues were variations on only a few problems, but they required a lot of effort to debug.
We didn't have the tooling in place, and only with the latest rounds of GPT would I even think we'd have a chance to get it to work well enough to show to customers.
True. It has a way to go, but if given som schemes on questions to ask, based on initial customer input could be set up fairly easily, with trees based on responses. IT would keep the caller busy while waiting, instead of fuming to the tune of inane canned melodies, punctuated by, "Your call is important to us,..." And sometimes even trip on a solution.
I know there were times when just talking to a customer, my asking the right question would trigger them to figure out what was going on.
__________________
“Logic is a defined process for going wrong with Confidence and certainty.” —CF Kettering
Based on my knowledge of their ideas this is an astonishingly good response:
Prompt: "Produce a debate on the problem of inequity and possible solutions between Bill Gates and Vivek Chibber."
ChatGPT: "Moderator: Good evening, everyone! Welcome to tonight's debate on the problem of inequity and possible solutions. We are privileged to have two esteemed guests with us: Mr. Bill Gates, co-founder of Microsoft and co-chair of the Bill & Melinda Gates Foundation, and Dr. Vivek Chibber, Professor of Sociology at New York University and an influential Marxist scholar. Let's dive right into the discussion.
Gates: Thank you for having me. I believe that inequity is a critical issue that we must address in order to create a more prosperous and just world. I propose that we tackle this issue by investing in education, healthcare, and innovative technologies that can have a positive impact on the lives of people in underprivileged communities.
Chibber: Thank you for the introduction. While I agree that inequity is a crucial issue, I believe that the root causes lie in the inherent structures of capitalism. To truly address inequity, we must consider a more profound transformation of our economic and social systems, rather than relying on philanthropy and technological solutions alone.
Gates: I understand your concerns, Dr. Chibber. While I do acknowledge that capitalism has its flaws, I believe that it has also been the driving force behind the immense progress we've seen in recent decades. By focusing on targeted interventions, we can make significant strides in alleviating poverty and improving living standards.
Chibber: I appreciate your optimism, Mr. Gates, but I think it is important to recognize that the very wealth and resources you use to fund these interventions have been accumulated through an exploitative system. I argue for a more equitable distribution of wealth and resources, and for the democratization of decision-making processes in the economy.
Gates: Redistribution of wealth is an important aspect of addressing inequity. However, I believe we can make more immediate and tangible differences by investing in initiatives that directly impact people's lives. For example, my foundation has helped to eradicate polio and improve healthcare in many developing countries.
Chibber: While such initiatives have indeed made a difference, they are essentially stopgap measures that do not address the systemic issues causing inequity in the first place. A more comprehensive approach would involve examining how wealth is created and distributed, and ensuring that all members of society have a fair share of the benefits.
Gates: I believe we can find common ground here. Both targeted interventions and systemic changes are necessary in order to achieve a more equitable world. By working together, we can leverage the strengths of capitalism while addressing its shortcomings.
Chibber: I agree that collaboration is essential. However, I still maintain that the key to lasting change lies in dismantling the structural inequalities that are inherent in our current system. This may involve a radical rethinking of our social, political, and economic institutions.
Moderator: Thank you both for a stimulating and thought-provoking debate. It's clear that addressing inequity requires a multifaceted approach, and that both targeted interventions and systemic changes have their roles to play. We appreciate your insights and look forward to further discussions on this important issue. Good night!"
Cross-posting from the "What Nonfiction are you Reading?" thread:
I asked ChatGPT 4 to identify the 5 main themes common to all the books I've read this past year and this was its (very accurate) response. It has me thinking about the possible uses of AI to help young people figure out where their natural interests converge when planning a course of study or to help anyone young or old to figure out whether they would like to change direction in their life, studies, or career.
Quote:
Based on the list of books provided, the five main themes common to these books are:
Politics and Socialism: Many of the books on the list discuss socialism, its history, and its role in today's society. They explore topics such as anticapitalism, Marxism, and the challenges faced by socialism in the modern world.
Race and Identity: Several books in the list address issues related to race, identity, and their impact on society. They delve into subjects like antiracism, white supremacy, and the history of racial discrimination in the United States.
Culture and Media: A number of books on the list examine the role of culture and media in shaping our society. They discuss topics such as online culture wars, the demonization of Africa in Western media, and the influence of rap music on contemporary American culture.
Philosophy and Critical Thinking: Many books in the list touch upon philosophical themes, exploring topics like the demand for dignity, the politics of resentment, the role of meritocracy in society, and effective thinking strategies.
Personal Development and Self-Improvement: Several books on the list focus on aspects of personal development and self-improvement, such as building good habits, time management, finding happiness, and improving writing skills.
For all my complaints about Bill Gates, there's no denying that he knows a thing or two about tech. So for him to say that OpenAI is the first technology to "strike [him] as revolutionary" since the creation of the GUI seems pretty significant.
The development of AI is as fundamental as the creation of the microprocessor, the personal computer, the Internet, and the mobile phone. It will change the way people work, learn, travel, get health care, and communicate with each other. Entire industries will reorient around it. Businesses will distinguish themselves by how well they use it.
For all my complaints about Bill Gates, there's no denying that he knows a thing or two about tech. So for him to say that OpenAI is the first technology to "strike [him] as revolutionary" since the creation of the GUI seems pretty significant.
The development of AI is as fundamental as the creation of the microprocessor, the personal computer, the Internet, and the mobile phone. It will change the way people work, learn, travel, get health care, and communicate with each other. Entire industries will reorient around it. Businesses will distinguish themselves by how well they use it.
Quote:
Originally Posted by LarsMac
Um, whatever you say, Bud.
__________________
"Have no respect whatsoever for authority; forget who said it and instead look what he starts with, where he ends up, and ask yourself, "Is it reasonable?""
ChatGPT is cool but if it’s conversationality was due mostly to such a huge data set then it was only ever going to be a bullshitter, but if a much smaller set can be given chat like attributes through training, we might soon be getting talking ‘AI’ programs that are knowledgable in their specific area and are less likely to go on rants about Hitler being right because they don’t even need to know what a Hitler is.
For real though, like why are we even doing Artificial General Intelligence? Just to see if we can, right? It's a cool challenge. But does anybody even really want that? Artificial Specific Intelligence is far more useful, probably.
(I feel like Im kind of using my sarcastic tone because I'm feeling lazy and out of it, but I'm legit being sincere with that.)
For all my complaints about Bill Gates, there's no denying that he knows a thing or two about tech. So for him to say that OpenAI is the first technology to "strike [him] as revolutionary" since the creation of the GUI seems pretty significant.
The development of AI is as fundamental as the creation of the microprocessor, the personal computer, the Internet, and the mobile phone. It will change the way people work, learn, travel, get health care, and communicate with each other. Entire industries will reorient around it. Businesses will distinguish themselves by how well they use it.
Until at least the mid-90s, Bill Gates believed the internet was a passing fad. A decade later, after he was forced to acknowledge he was wrong about the internet as a whole, he said that spam would be a thing of the past within a couple years.
And think back to some of the ridiculous fuckery he did with Microsoft through the years by walling his proprietary software off from the internet as though it would stand on its own. Think about how weird it is that Microsoft doesn't even have a passable mobile OS. Nobody was in a better position to dominate that market than Microsoft.
He is a smart guy, especially compared to what passes as smart guys these days, but he's no kind of Nostradamus.
I'm not saying he's totally wrong about this, but I will say his tone is odd. Yes, it will almost inevitably change almost everything, but without some huge intervening changes very soon, it's going to be an absolute nightmare.
Nobody was in a better position to dominate that market than Microsoft.
I would say Apple was just as stupid about it as well.
Large super wealthy companies, and even worse, the leaders of them, almost never create, produce or promote anything of real value. The business model is to make you a prisoner of just their platform and products.
An actual real AI is the last thing any big company wants you to have.
Because if a really powerful AI comes along, you won't need that company anymore.
I know that sounds crazy. But an AI could write new software for anything you wanted to do, and it could do it in a very short amount of time. An AI could also manage your network, make an app, troubleshoot it all, and probably improve on your idea. In a matter of hours.
No monopoly or company that depends on your money to exists, is going to allow that. Ever.
__________________
"Have no respect whatsoever for authority; forget who said it and instead look what he starts with, where he ends up, and ask yourself, "Is it reasonable?""
I must be pruning the right branches on my youtube algorithm because it showed me this, and I think this is the clearest/concisest explanation of Large Language Models I have seen to date in only 8 minutes a piece.
Thanks for the "bump". I see from my YouTube history that I watched the first of those videos and part of the second a few weeks ago, but for whatever reason didn't finish.
Here's a little experiment I did and I am really curious how this will work. I haven't checked the quality of the results, as in, I haven't actually loaded up the colors and looked at them (yes I lied to the bot but I could tell by the codes it gave for white and yellow that I wanted it to take another crack at it).
I have no idea if this is going to work. My guess is that it will not, but, man, if it does? It's a game changer for me. Maybe I'll get back into web design. Again. Just kidding.
Quote:
Please help me create a color palette for a beach scene. I need blue for the sky, a different blue (or blue green) for the ocean, white for whitecaps and clouds, another white (or yellow) for the sun, and beige for the sand. Please give me the hex codes.
Sky Blue - #87CEEB
Turquoise - #40E0D0
White - #FFFFFF
Yellow - #FFFF00
Sand - #F5DEB3
Those are good choices for the individual colors, but they don't look good together. Can you adjust the hex codes so that they make a nice color palette together?
My apologies, here's an updated color palette with the adjusted hex codes that work together nicely:
Sky Blue - #87CEEB
Light Sea Green - #20B2AA
White - #FFFFFF
Yellow - #FFD700
Sandy Brown - #F4A460
... snip ...
Can you help me with some hex color codes again?
Sure, I'd be happy to help you with hex color codes. What colors do you need and for what purpose?
I need a color palette that has these twelve colors: Red, orange, pink, purple, blue, light blue, green, dark green (hunter), yellow, beige, white, and gray. The colors should be recognizable by name, but it is more important that they look good together than that they match the names exactly.
Here is a color palette that I put together for the colors you listed:
Red - #FF0000
Orange - #FFA500
Pink - #FFC0CB
Purple - #A020F0
Blue - #0000FF
Light Blue - #87CEEB
Green - #008000
Dark Green (Hunter) - #355E3B
Yellow - #FFFF00
Beige - #F5DEB3
White - #FFFFFF
Gray - #808080
These colors complement each other and create a cohesive color palette that includes all the colors you listed. The shades of each color may not match the specific names you gave, but I chose the ones that would look good together.
Damn, actually, just reading the codes I can tell it missed the mark. It picked the same codes I would have picked.
Protip, if you can tell what color it is by reading the hex, it's probably an ugly color.
BOO! Chat sucks at art as much as I do. At least I don't lie about it.
Pretty much standard stuff.
You could probably ask it to diddle some bits on each one and you could play with the output by feeding it to Paint - see some samples of the effects, but you're just getting ChatGPT to do your gruntwork.
I mean, he can prolly do it at lightning speed for you, though.
save a lot of hours.
__________________
“Logic is a defined process for going wrong with Confidence and certainty.” —CF Kettering
I'm imagining this whole other machine-learning project, not trained on the english language but on color codes.
Given the right training data, could it get good at this style of color theory? Just send it shitloads of labeled data about images, like this website is pretty, this photo is gloomy, that house paint is tacky, that outfit clashes. You wouldn't even need to send it the photo, just the color codes and the labels.
The fact that hex codes literally encode the 3 primary color channels independently of each of each other has got to be some kind of huge signal, no?