OpenAI Discussion

General
#1

It is the biggest OpenAI product launch since the original ChatGPT debut.

What it is:
A business tier for the AI chatbot available starting today.

How it rates against the consumer tier:
Access to GPT-4 with no usage caps, up to 2X faster performance and API credits.

What it does:
“…allow clients to input company data to train and customize ChatGPT for their own industries and use cases…”

How it came to be:
Developed together with more than 20 companies of differing sizes and sectors. Beta users included Block, Canva and Estée Lauder.

Pricing:
Bespoke for each company’s use case and size.

Other tier:
There are plans to launch “ChatGPT Business” for smaller teams.

Does it exploit your data?:
“OpenAI noted in a blog post that ‘We do not train on your business data or conversations, and our models don’t learn from your usage’, adding that clients’ conversation data would be encrypted both at transit and at rest.:”

Full info and source for this post:

3 Likes
Artificial Wisdom
#2

This is fantastic! It was always so awkward to have to train your replacement in person.

#3

Funny @Nancy_Muellerhof

I think the way we have to look at is more in a positive way.

When the automobile was invented, thousands of horse saddle makers were no longer needed. They were brilliant at working with leather so they were hired by the car companies to sew the seats for the cars. A lower salary job with less satisfaction, but a job at least.

Ok I made that up. Maybe it happened I don’t know. You get the point.

#4

“It is not the strongest of the species that survives, nor the most intelligent that survives. It is the one that is most adaptable to change.”

#5

Right @MarkOlsen new technologies have changed the workplace for centuries. However the speed at which it is happening is increasing. People born around 1970 have been through the birth of computing, the birth of the internet, two severe economic crisis, globalization, and now AI. Never has a generation experienced change like Generation X.

A short but interesting treatise on the subject:

#6

I’m not as optimistic that people can simply adapt to the reality of AI in the same way that previous transformative tech has been adapted to. Until now, even a task as basic as picking strawberries has had to be done by human beings. There was simply no technology that could reliably look at the plant and know what to grab, and what not to.

AI is transformative in an unprecedented way. The one thing we humans have always had going for us is that computers did not have judgement and creativity. I make my living in a creative field. The product that AI creates, which would compete with me, is still pretty mediocre. But it is good enough for a significant percentage of applications. It seems that creative fields will mostly only survive at the very high end for clients with very specific and exacting requirements.

1 Like
#7

I ask Open AI ChatGTP to answer your post. Be assured that it hopes for a future where AI and human creativity coexist in harmony. It seems like a speech from someone who has already won the battle; see below:

I understand your concerns about the transformative impact of AI and the challenges it poses, particularly in creative fields. The comparison between adapting to previous transformative technologies and adapting to AI is indeed a thought-provoking one.

Historically, we’ve witnessed societies adapting to various technological advancements, from the Industrial Revolution to the rise of computers and the internet. Each wave of innovation brought its own set of challenges and adjustments, but ultimately, humans found ways to integrate these technologies into their lives and create new opportunities.

However, AI does present unique challenges due to its unprecedented ability to simulate human-like tasks, such as decision-making, language processing, and even creative tasks like generating content. Your example of AI-generated content in creative fields reflects the current state of AI, where it can produce outputs that are functional and often good enough for certain applications, but might lack the depth, nuance, and originality that humans can offer.

It’s important to acknowledge that AI’s capabilities are evolving rapidly, but there are aspects of creativity, intuition, and emotional understanding that remain deeply rooted in human experience. While AI-generated content might suffice for specific tasks or industries, there’s a strong argument to be made that true innovation and breakthroughs will continue to come from the human touch.

In creative fields, the potential impact of AI could indeed lead to a shift in how work is valued and where creativity truly shines. The high-end, specialized requirements you mentioned are likely to remain a stronghold for human creativity, as clients seek unique and deeply personalized solutions that only human creators can provide. The coexistence of AI-generated content and human creativity could lead to collaborations that leverage the strengths of both sides.

Adapting to AI’s transformative power might not be as straightforward as previous technological shifts, but it’s a challenge that can be met with a combination of embracing new tools, honing uniquely human skills, and redefining the value of creativity in a changing landscape. The journey ahead will undoubtedly involve navigating uncharted territory, but the human capacity for innovation and adaptation remains a constant that gives us hope for a future where AI and human creativity coexist in harmony.

#8

When I read this I hear the calm voice of HAL from 2001 Space Odyssey.

#9

As a counterpoint, a quote from John Oliver that quite resonates with my experience of ChatGPT and current AI tools:

“The problem with AI right now isn’t that it’s smart. It’s that it’s stupid in ways that we can’t always predict. Which is a real problem, because we’re increasingly using AI in all sorts of consequential ways.”

2 Likes
#10

Amazon Web Services and NVIDIA are collaborating on a scalable AI product that will presumably compete with ChatGPT Enterprise:

#11

A group of prominent novelists are suing OpenAI.

“The authors claim their books were used to train the ChatGPT Bot without permission or compensation”

“…can spit out derivative works that based on, mimics, summarizes, or paraphrases (the author’s work) and harms the market for them”

I use ChatGPT for recipes all the time. Does it invent them? I asked ChatGPT if it uses other people’s work, and it gave an ambiguous answer and claimed it would credit any sources if it did. I’ve never seen it credit any source however.

#12

Hey everyone, thought I would pose a question:

What would you have to see or experience to reach a point where you feel that A.I. has crossed the line from being a helpful tool, to being a nuisance, or even a threat?

2 Likes
#13

I think the concerning part is not what AI does. What would concern me is how much we allow it to do. My redline would be seeing people sit at a bar and have a conversation with a robot bartender that makes perfect drinks every time.

That would be the end of humanity.

#14

I believe the annoyance we associate with AI will diminish as it evolves. Chatbots, for example, will soon respond to questions with greater accuracy, without ‘breaking’ as frequently.

And I’m actually okay with the novelty of robot bartenders provided the patrons are still returning to their date or group of friends after conversing with the machine and picking up their perfect martini. :cocktail: Where my concern begins to grow, (very much aligned with this bartender example, but extrapolating further) is with the rising popularity of AI companions such as Replika.

I think this is particularly significant for younger generations learning to form strong bonds, life-long friendships, and to experience heartbreak and rejection, as well as navigate the turbulent waters that relationships often bring. The ‘always there,’ ‘always listening’ asymmetrical nature of AI companions might feel like an easier path, circumventing the difficulties of relationship building altogether. Recent advancements in LLMs (large language models) that provide lifelike interactions could make these virtual companions even more appealing. However, the convenience they offer may also increase emotional isolation and unrealistic expectations, potentially leading to a loss of authentic human connection and broader macro effects such as a further decrease in demographic trends.

AI companions are already a trend in countries such as Japan and Korea. Korea’s world’s lowest fertility rate (0.78 children), with Japan not far behind, underscores a demographic challenge. Economic uncertainty, cost of living, and decreased interest in marriage and children have played a part, but an increase in AI companions may further exacerbate this issue.

While caution surrounds other areas of AI development, like military use and misinformation (concerns I share), the potential problem of AI replacing genuine relationships is largely overlooked. Seems to me like a problem that could sneak up on us over the next decade, and yet feel inevitable in hindsight.

1 Like
#15

Maybe you have seen the Spike Jonze movie “Her”?

It is a social satire about a guy who falls in love with an AI companion. When the movie came out 9 years ago it was kind of funny. Seeing this guy having emotional conversations and a real relationship with his computer was ridiculous. It was absurd drama/comedy. It was a cautionary tale about what could happen in 50 years or so.

And now it is real? I had no idea that “AI companions” were a thing. Young people actually forming emotional bonds with an algorithm of some kind that learns how to constantly tell them what they want to hear. It is beyond comprehension but it looks like the genie is out of the bottle so to speak.

1 Like
#16

@IanT The strange thing about watching that film for me was falling in love with the AI character myself! It messes with you mind a little.

My teenage son is currently experiencing the woes of puppy love, and often talks about how he just want to be alone. His isolation, and as much time as he spends on his phone watching one short clip after the other, it is actually feasible to me that he could find comfort in conversations with an AI companion. I can’t believe I’m writing this.

I proactively break the spell of technology with him every day, engaging him in conversation, asking him questions that he has to think about, and getting him out of the house and involved with things. But if didn’t have that…

@Jo-CKC-Studio You mentioned Replika. When I landed on the home page, my first thought was that it was an article in “The Onion” but the realtity hit quickly. This is real and it is happening. Perhaps it is progress, perhaps is inevitable, perhaps every aging generation looks on in dismay as technology replaces human activities. However seeing Replika for the first time was the most Orwellian experience I think I’ve ever encountered. And I used to work in Moscow!

1 Like
#17

There’s another thread on this forum discussing the new ChatGPT product for enterprises. This raises a new question about AI red lines, originally posted by @Nancy_Muellerhof. How willing will team members be to train the AI to ultimately replace them. Will they sabotage the AI with incorrect input?

#18

That is a great question. Will employees sabotage AI as they are “training” it? Will they train it to make mistakes to avoid being replaced? Wow.

If you run an enterprise, it is clear that it would be wise to take into account the genuine human cost in terms of anxiety and all of its ripple effects, that the introduction of AI into your workflow will cause among your employees.

Who wouldn’t be scared and a bit resentful of training a machine to replace them?

#19

Technology affecting jobs isn’t new. Like past innovations, AI will bring labor market shifts. But as AI enters knowledge work, upskilling certainly will become crucial to avoid redundancy. There is a large contingent of ‘blue collar’ work within the ‘white collar’ industry, such as call centers, support lines, administrative roles, processing/data entry, and basic analysis and bookkeeping. However, for many knowledge industry workers, these AI tools can help reduce repetitive areas of their work so they can focus on more interesting and creative elements, increasing productivity. If the job is fully repetitive or ‘programmable,’ AI may fully replace it. But for most, AI is another powerful toolkit enhancing productivity alongside human strengths like creativity.

Regarding disgruntled employees potentially sabotaging knowledge bases while training them, if misuse starts happening, I’m fairly confident that the company will make fixes to stop it. It’s unlikely though. Whenever you hear “this call will be recorded for training purposes,” that’s what’s happening. I remember working with a client who was putting these big data collection systems in place back in 2014 (incoming calls translated to text, AI categorizing and tagging data so the company can work out who and why people call, to more efficiently route calling systems and work out resourcing). Data collection for ‘training’ has been going on for some time now, so if sabotage was going to happen, it likely would have surfaced earlier. The technology is just more widely talked about now

#20

I also enjoyed the film “Her” - great characters and creative storytelling. I particularly liked that it avoided dystopian sensationalism. When I first watched it, the premise still felt far-fetched, yet we’ve already closed much of that gap. While parts remain fictional, humanlike companions are certainly here now. We are living in a somewhat strange world.

Personally, I think that for many (especially younger generations), making meaningful friendships online is likely no different from traditional socializing - catching up with friends in real life. Online interactions might seem strange from the outside, but for many it’s quite social, just in a different way. Still, getting some fresh air, and proactively participating in ‘real life’ family/social/community events are important for all forms of health, so it shouldn’t be an either/or.

To me, the crucial differentiator is whether those online bonds/friendships, however genuine, are with fellow humans. There’s an infinitesimal gap between befriending people online or off, yet a vast distance remains between human friendships and even the most socially adept AI - it is certainly not the same thing.

1 Like