Your key to devising GPTs with the unveiled OpenAI capability and making erstwhile money from your … [+] efforts.
gettyA golden opportunity has been unveiled by AI maker OpenAI for those of you who want to turn your use of ChatGPT and GPT-4 into a money-making activity.
Yes, that’s right, serious dough could be had. In short, a new feature known as GPTs gives you a solid chance to go from simply being a hobbyist or fun-filled user of generative AI into someone who garners nice warm cash for your online labors. Plus, whether you want to do so on a heady full-time basis or just as a part-time side hustle is entirely up to you.
In today’s column, I will lay out what this new announcement by OpenAI foretells and indicate how you can leverage the big break that has arisen. It is then up to you to decide what to do next.
Let’s jump in.
Time is of the essence.
Able To Devise Generative AI Miniature Applets Known As GPTs
Okay, so OpenAI refers to this exciting new feature as the ability to craft GPTs.
Here’s how they officially describe this:
“We’re rolling out custom versions of ChatGPT that you can create for a specific purpose—called GPTs. GPTs are a new way for anyone to create a tailored version of ChatGPT to be more helpful in their daily life, at specific tasks, at work, or home—and then share that creation with others” (as quoted from the OpenAI website blog entitled “Introducing GPTs” posted online on November 6, 2023).
The way to think of these GPTs is that they are essentially customized versions or miniatures of ChatGPT.
Step by step, it goes like this.
You go into ChatGPT and do some special setup based on either something that you know or something that you believe others want to know. You save the instance. Your newly devised GPT then becomes a kind of miniature applet. Think of this as akin to becoming a canned kind of app that others can run (within the ChatGPT environment). All you had to do was enter various prompts to guide ChatGPT to get it going in some particular direction of action, and then you perform administrative chores that turn the instance into a runnable applet that others can make use of inside the ChatGPT environment.
The upside to all of this is that there are already 100 million active users of OpenAI’s products and thus there is a built-in market ready to make use of your devised concoction. OpenAI is going to have an online store wherein chosen GPTs will be listed for use. Whenever someone opts to use your devised GPT, you and OpenAI will get a share of the revenue produced. OpenAI takes care of the online housing of your GPT and all of the behind-the-scenes money collection. You get yourself a payoff commensurate with how often your GPT is used and based on other factors (they’ve not yet articulated the payment scheme to be used).
You do not need programming skills to get things underway. All you need is the essentials of using generative AI. The rest is pretty straightforward to undertake.
As an aside, you might have heard of a movement in the software field toward no-code software development. The notion is to allow non-programmers to craft programs or apps, without having to learn how to write code. These GPTs are just like this. If you are already familiar with the simple acts of entering prompts into ChatGPT or any generative AI, you know what you need to know. Forget about those software programming classes that have been languishing on your tiresome To-Do list.
OpenAI depicts the situation this way: “Anyone can easily build their own GPT—no coding is required. You can make them for yourself, just for your company’s internal use, or for everyone. Creating one is as easy as starting a conversation, giving it instructions and extra knowledge, and picking what it can do, like searching the web, making images, or analyzing data” (ibid).
I know that sounds almost too easy.
One important caveat is that if you do proceed with devising one or more of these income-producing GPTs, please realize that you better be smart about doing so. You need more than wild dumb luck to be successful. I would strongly advise for example that as a minimum, you learn about and make use of proper prompting strategies for generative AI, as I have covered extensively in my ongoing series, see the link here to get started. You will need to write suitable and erstwhile prompts to be able to devise a GPT that people are going to want to use. Prompt engineering will be your best and closest friend.
Here’s the plan of what I will cover next. I will provide explicit guidance to get you going. I will also share with you the gotchas to be mindful of. Do your best to aim properly and avoid falling into numerous abysmal downtrodden traps.
Things You Know Or Things People Want To Know
A sensible way to conceive of coming up with what your GPT will do is to consider these three major considerations:
(1) Things you know
(2) Things people want to know
(3) Things you know and that handily dovetails into things that people want to know.
Let’s start with the first point, namely focusing on an approach involving things that you know.
Are you someone who is versed in a particular field of endeavor or a specialized skillset? If so, you might want to consider crafting a GPT based on your seasoned base of knowledge and lived experiences.
That being said, you don’t have to have some kind of lofty set of skills. I realize your first impression might be that only if you are a rocket scientist would anyone want to use a GPT that you devised.
Not so.
Suppose that you happen to play Minecraft extensively and have fine-tuned your playing skills regarding the popular online video game. You could go into ChatGPT and craft an instance that encompasses prompts telling ChatGPT how to advise someone who really wants to optimize their Minecraft strategies and tactics. After initially doing so, you then do some double-checking in ChatGPT to see how well ChatGPT has garnered what you’ve stipulated. Assuming you then believe that the result is something others might find useful when playing the game, you establish your newly minted GPT accordingly.
The vaunted hope is that of the 100 million or so users of OpenAI’s products, some will want to make use of your GPT. Your audience would presumably be people who are either already Minecraft aficionados and want to up their gaming talents or those who are new to playing Minecraft and want to get up-to-speed with a handy Minecraft-advising generative AI applet.
I’ll give you another example to get the wheels turning in your head.
A plumber has repeatedly over the years learned all sorts of ways to deal with a jammed-up sink. Most of the time, a homeowner could fix the jam themselves but they don’t have a clue as to how to do so. People might sometimes do an Internet search on how to fix a clogged sink. The problem there is that the information posted can be quite generic and not especially helpful in a given dire situation.
A GPT could be crafted that explains how to fix a clogged sink. Furthermore, since it is using generative AI, the GPT would fluently interact and ask the person what kind of sink they have, how it got clogged, what tools they have on hand, and so on. The advice about fixing the clog would be honed to the circumstances. This is a lot more alluring than just looking at some generic posting on the Internet.
We can up the ante on the plumbing example. Other new features are coming online such as being able to capture a camera image and have the generative AI examine the image. This is big. A person takes a picture of their clogged sink and the plumbing under the sink. If the generative AI is vision processing capable, it can examine the pictures and hone the advice about what to do. You went from the person having to type in a description of their sink and instead have the AI analyze the sink predicament via snapped pictures.
I trust that you get my drift as to considering leveraging the things that you know.
Next, let’s discuss things that people want to know.
It seems pretty clear these days that people are for example craving relationship advice. Tons of social media influencers have come out of nowhere to become high-earning relationship-advising influencers. Some of those are bona fide experts and have specialized training or garnered notable college degrees in the subject matter. Others are talking based on not much more than the school of hard knocks after having had a slew of busted relationships. It seems nearly anyone can be a relationship advisor nowadays.
The crux herein is that you might diligently and intelligently seek to identify something that people want to know. Once you’ve come up with a potential interest that requires advisement or guidance, you prepare by trying to come up to speed on the topic. Doing so sufficiently to be able to go into ChatGPT and get it headed in the direction of that realm or domain of advisement. Voila, you then craft a GPT on it.
Finally, consider the third option.
I earlier noted that you can do a GPT on the things you know or do the GPT on things that people want to know. Ideally, you want to attain both paths at the same time.
Here’s what I mean.
If you devise a GPT on something that you comfortably know, but the topic is not something that people particularly care about, you will have a GPT that few will opt to use. This means that you won’t make much money at all on the GPT. I am not trying to discourage you from this. Your GPT still might be helpful to people and quite worthwhile devising. Pursue your passions. Help the world.
On the other hand, if you devise a GPT based on something that people seem to want to know about, the chances are that you might get lots of usage. The usage in turn will bring in the bacon. The thing is, if you aren’t especially versed in the topic, one wonders how good the GPT is going to be. A thin GPT that doesn’t have much depth might not get usage. Your attempt to skirt past having expertise in something could land on a dud of a GPT.
I assume you got the picture.
Your best bet would be to be lucky enough to know something well that people also want to know about. I dare say that isn’t necessarily the case much of the time. Think about it. You never know what it is that people might want to know about, even if the things you know are not currently popular or touted in the media. There is a reasonable chance of having a sleeper topic that sparks interest and gets viral usage suddenly and unexpectedly.
Go for it.
You Are Going To Compete In A Darwinian GPT Competition
So far, you might feel pretty upbeat on this new GPT capability. I’m glad. Please remain upbeat as I shift gears somewhat.
I noted that there are caveats to be aware of. Here we go, trigger alert.
Shocker: The odds are that a lot of people are going to pursue devising GPTs.
You aren’t the only one doing so.
Furthermore, and this is the potential energy-sapping spirit-crushing difficulty, you will most likely strenuously desire to have OpenAI approve posting your GPT into their store. To clarify, you don’t have to get posted for people to use your GPT. You can otherwise still make it publicly available. The issue here is that you’ll also need to essentially market your GPT on your own. I’m sure this will work out sometimes. The smoother angle would be to get your GPT into the official OpenAI GPT Store, which also will give your GPT an aura of grandiose respectability.
Here’s what OpenAI has to say about the GPT Store:
“Starting today, you can create GPTs and share them publicly. Later this month, we’re launching the GPT Store, featuring creations by verified builders. Once in the store, GPTs become searchable and may climb the leaderboards. We will also spotlight the most useful and delightful GPTs we come across in categories like productivity, education, and ‘just for fun’. In the coming months, you’ll also be able to earn money based on how many people are using your GPT” (ibid).
I wager that getting your GPT into the GPT Store will be a tough road to hoe.
If your GPT is akin to a thousand others, why should your GPT be included versus a similar one?
Once a particular niche gets occupied, it might be tough to make the case that your GPT is better than an incumbent. One would also guess that the selection process for getting into the GPT Store will partially aim to not have numerous apparent duplicates that all do the same thing. This would seem imprudent and cause people to doubt the efficacy of the GPT Store.
My suggestion is as follows:
Keep your eye on the GPT Store to see what is being included.
Try to aim your GPT at a realm or niche that isn’t yet included.
Make the best GPT you can in your chosen realm or niche so that it can stand out.
Prepare to explain or showcase why your GPT should win the Darwinian thinning process.
If you don’t get into the GPT Store, don’t give up, market the heck out of your GPT anyway.
Learn from the above and possibly come up with another GPT on something else.
Though it is likely the GPT Store will be the most productive venue for your GPT, this is not the end all. You could become famous for a GPT that is not in the GPT Store that you’ve made publicly available and somehow miraculously catches the public eye. You might then parlay that fame into other money-making possibilities.
Please keep that happy face scenario in mind during any dark days associated with your GPT ideas and devising.
GPTs Upcoming Examples To Be Explored
Some examples of GPTs are going to be posted by OpenAI.
Assuming you already have access to ChatGPT, you might want to take a look at them once they get posted. Doing so could give you some ideas of how they work and could inspire you toward devising a GPT of your own.
The other aspect is that seeing the examples might make you realize that there is nothing especially earth-shattering about them. Allow me to expound. It is tempting to assume that only something star-studded would be worthy of putting together as a GPT. That’s not what is going to occur, at least at first.
My best guess is that at first impulse we will have all manner of everyday style GPTs. After the initial madcap dash to devise GPTs, bombarding the world with so-called AI “agents” (a terminology often associated with these kinds of AI applets), efforts will extend toward more out-of-the-ordinary arenas. The easy topics will get swamped in the initial wave. The harder or more outlier topics will remain as viable untouched terrain.
Here is a quick taste of some examples that are purportedly going to be listed on the OpenAI blog:
“Game Time: I can quickly explain board games or card games to players of any age. Let the games begin!”
“Creative Writing Coach: I’m eager to read your work and give you feedback to improve your skills.”
“The Negotiator: I’ll help you advocate for yourself and get better outcomes. Become a great negotiator.”
“Laundry Buddy: Ask me anything about stains, settings, sorting and everything laundry.”
“Sous Chef: I’ll give you recipes based on the foods you love and ingredients you have.”
“Math Mentor: I help parents help their kids with math. Need a 9 p.m. refresher on geometry proofs? I’m here for you.”
“Sticker Whiz: I’ll help turn your wildest dreams into die-cut stickers, shipped right to your door.”
“Tech Support Advisor: From setting up a printer to troubleshooting a device, I’m here to help you step-by-step.”
Those seem rather vanilla if you know what I mean.
You can right away see some examples depending upon your OpenAI account: “Example GPTs are available today for ChatGPT Plus and Enterprise users to try out including Canva and Zapier AI Actions. We plan to offer GPTs to more users soon” (ibid).
You Must Follow The Rules Or Else Be Booted Out
I’ve got some rough news, though I think you will see the sensible logic involved.
First, most users of ChatGPT and GPT-4 (and likewise with other generative AI apps) seem to be oblivious to the aspect that they agreed to a licensing agreement associated with the AI when they signed up to use the generative AI. I have discussed at length that the AI makers usually list an indication of all kinds of banned uses that you aren’t allowed to undertake with their generative AI, see the link here.
I bring this up because you ought to examine the banned list before starting to devise your GPT.
No sense in making a GPT that right off the bat will get turned down by OpenAI as a banned use. Even if you don’t opt to gain entry into the GPT Store, if you publicly make your GPT available, the chances are that one way or another OpenAI might get wind of it. If your GPT is doing things that are on the banned list, the AI maker can close down your GPT, plus other consequences are spelled out in their usage policies.
Here’s what OpenAI officially says on GPTs and the usage policies:
“We’ve set up new systems to help review GPTs against our usage policies. These systems stack on top of our existing mitigations and aim to prevent users from sharing harmful GPTs, including those that involve fraudulent activity, hateful content, or adult themes. We’ve also taken steps to build user trust by allowing builders to verify their identity. We’ll continue to monitor and learn how people use GPTs and update and strengthen our safety mitigations. If you have concerns with a specific GPT, you can also use our reporting feature on the GPT shared page to notify our team” (ibid).
As you can see, the AI maker is increasing their vigilance. Before, people might have been doing banned uses but presumably only they saw their instance anyway. Now, with being able to make a GPT publicly available, the onus on the AI maker has been raised. You might also have noticed that a “snitch” line can be used to report concerns about a specific GPT.
I should state abundantly and loudly that the usage policies are there for very good reasons. Do not try to circumvent them. Do not try to be clever and use tomfoolery to outmaneuver the rules. Readers of my column know that I harp time and again on the importance of having Responsible AI, see for example the link here and the link here.
Be responsible and make GPTs that are good for the world.
Do not descend into evil uses of AI.
Privacy And The Matter Of Those GPTs
If you took a reflective moment to think deeply about the GPTs, you might come to a thought that might be unnerving.
Here it is.
Suppose that you use a GPT that someone has devised. You do so and enter private information into the GPT. Does the person who devised the GPT get to see your private information? I dare say that if that happened, people would avoid using GPTs like the plague. You would potentially be exposing private aspects and not necessarily know who is going to get it and what they might do with it.
The answer is somewhat complicated.
Let’s see what OpenAI says in their blog posting:
“As always, you are in control of your data with ChatGPT. Your chats with GPTs are not shared with builders. If a GPT uses third-party APIs, you choose whether data can be sent to that API. When builders customize their own GPT with actions or knowledge, the builder can choose if user chats with that GPT can be used to improve and train our models. These choices build upon the existing privacy controls users have, including the option to opt your entire account out of model training” (ibid).
There is a lot there to unpack. In brief, those who opt to use a GPT will need to understand their options and realize how their settings come into play in this matter.
For example, I have noted in my column that depending upon the privacy control settings you’ve chosen, you might be allowing the AI maker to essentially have access to the prompts that you enter into ChatGPT, see my analysis at the link here. Those prompts can potentially be seen by AI developers or AI testers of the AI maker and can possibly be utilized for further data training of the generative AI.
My rule of thumb that I tell people is you should normally avoid entering private information into any generative AI. You are safer to avert doing so. Sadly, I’ve seen people using generative AI that pour their hearts out and provide all manner of private information that they ought to not be revealing. It is easy to fall into this trap because the generative AI seems so friendly and inviting.
Just wanted to make sure this notable point is on your radar.
Liability Along With Your Potential Legal Exposure Due To Indemnification Too
I’ll make this a quick point.
In my prior columns, I have forewarned that your use of generative AI could legally get you in trouble if you devise a generative AI app that provides advice, and the advice is perceived by someone as having caused them harm, see the link here. You see, a person might use your generative AI app and then turn around and sue you. They are going to legally attempt to argue that your generative AI app gave them foul advice and you ought to compensate them for any damages and adverse repercussions thereof.
This hasn’t particularly arisen since most people have been merely using generative AI and not building apps based on generative AI. There isn’t that much of a chance of these kinds of beguiling legal situations arising.
The GPTs will provide ample fodder for such possibilities.
A can of legal worms is about to be opened.
Consider this (I’ll revisit an earlier example). A plumber devises a GPT. Someone uses the GPT to aid in fixing their clogged sink. The GPT tells the person to take off the lower pipe. The person abides by the advice. Water floods into their kitchen and completely destroys their expensive kitchen floor, plus wipe out other items in the kitchen. The person fervently believes that the GPT gave them bad advice.
Who will this upset person decide to sue?
You might be thinking it makes no sense to sue since the person was the one who took off the pipe. They ought to be responsible for their own actions. Blaming the GPT seems zany. Well, all I can tell you is that our world seems to have gone zany when it comes to launching lawsuits. Maybe the person will have the lawsuit tossed or maybe they will win. We will have to see how these matters evolve.
Back to the question of whom to sue.
The plumber that made the GPT seems like a likely target to sue. Of course, the amount of money that might be collected from an everyday plumber is probably not much money (despite those seemingly astronomical prices to have a plumber come to your house). So, the alleged “harmed” person will indubitably try to sue the AI maker (grand holder of big bucks).
This is where a twist comes into the muddied matter. I’ve pointed out in my columns that as part of the licensing agreement, you usually agree to indemnify the AI maker for lawsuits against them for something that you did, see my analysis at the link here.
Go back to the plight of the plumber. The alleged “harmed” person sues the plumber and sues the AI maker. The AI maker invokes the indemnification clause and insists that the plumber needs to pay all the legal costs being borne by the AI maker. Yikes!
Generally, make sure to consult with your attorney when devising a GPT that might get you into legal straits. A bottom-line consideration is that we are going to enter into a realm of untested legal complications with the advent of the GPTs. You might not ever have any legal issues that arise. Or you might become a poster child precedent-setting court case.
I’ve predicted that attorneys are gradually and inevitably going to be expanding their legal services to encompass matters intertwining AI and the law, see my coverage at the link here and the link here. That day might be sooner than it seems.
Going With All The Bells And Whistles For Your GPT
The preponderance of people who will be devising GPTs will be ordinary folks who aren’t versed in heads-down programming or other arcane software development skills. But, if you are a sharp-as-a-tack software developer, you can use your savvy skills to make GPTs into something far beyond what an ordinary person might craft.
Feast your developer eyes on these announcements by OpenAI:
“In addition to using our built-in capabilities, you can also define custom actions by making one or more APIs available to the GPT. Like plugins, actions allow GPTs to integrate external data or interact with the real world. Connect GPTs to databases, plug them into emails, or make them your shopping assistant. For example, you could integrate a travel listings database, connect a user’s email inbox, or facilitate e-commerce orders” (ibid).
“The design of actions builds upon insights from our plugins beta, granting developers greater control over the model and how their APIs are called. Migrating from the plugins beta is easy with the ability to use your existing plugin manifest to define actions for your GPT” (ibid).
The takeaway is these three avenues are going to happen:
(1) Devise a GPT as an ordinary user
(2) Devise a GPT as a high-end developer
(3) Pair up an ordinary user with a high-end developer to devise a GPT
First, we are going to have gobs and gobs of ordinary users that will create GPTs. Meanwhile, we will have some amount of high-end developers that will also devise GPTs. Both camps are sensible and we can expect good things from each respective camp.
I would also anticipate that we will see teams formed. An ordinary user who is a specialist or skilled in some domain will pair up with a high-end developer who can make the devised GPT sing and soar. This increases the chances of the GPT rising adroitly on a heads and shoulders basis above the other competing GPTs. The real question will be how the two of you decide to split the proceeds from the money that the GPT makes. I recommend you get that figured out beforehand. Waiting until after the fact is usually irksome and just causes money-sapping legal battles.
Anyway, if you pair up, do so mindfully and with aplomb.
Companies Going Internal With GPTs Is Also Big Business
The above depiction of GPTs has focused on an outwardly expected use of GPTs, namely that you bring your GPT to the marketplace and seek globally recognized fame and fortune.
There is another route to be considered.
Suppose a company is using ChatGPT on an enterprise basis. The firm might want to devise GPTs that are solely for internal use. Maybe devising a GPT for the finance department would be helpful to the financial affairs of the firm. Perhaps the marketing department could use a GPT that aids them in their efforts. And so on.
Here’s what OpenAI says about the internal uses of GPTs:
“Since we launched ChatGPT Enterprise a few months ago, early customers have expressed the desire for even more customization that aligns with their business. GPTs answer this call by allowing you to create versions of ChatGPT for specific use cases, departments, or proprietary datasets. Early customers like Amgen, Bain, and Square are already leveraging internal GPTs to do things like craft marketing materials embodying their brand, aid support staff with answering customer questions, or help new software engineers with onboarding” (ibid).
“You can now empower users inside your company to design internal-only GPTs without code and securely publish them to your workspace. The admin console lets you choose how GPTs are shared and whether external GPTs may be used inside your business. Like all usage on ChatGPT Enterprise, we do not use your conversations with GPTs to improve our models” (ibid).
I bring this up for several reasons.
If you work inside of a company and the firm is using ChatGPT on an enterprise basis, you might want to see if you can get added to an internal effort of crafting GPTs. It would be a great experience for you. You can easily add this as a hot-button item on your resume. Plus, one hopes that it might make your work efforts more invigorating and engaging.
For those of you who aren’t in a company that is using ChatGPT Enterprise, but if you are someone who is a high-end developer, you might consider devising GPTs that you could potentially sell or license to firms that do have the product. Make sure to consult with your legal beagle about how to do this without garnering any legal concerns.
Conclusion
Insiders realize that the capability to customize or tailor generative AI is something that has already been promulgated by a gaggle of startups and tech firms. In that sense, many of those providers or vendors are feeling a bit perturbed that they already made headway in this sphere, and yet many don’t seem to realize that (in their eyes) this is nothing more than a reinventing of the wheel.
I don’t want to seem harsh, but the 600-pound gorilla has now spoken, and you have to learn how to live with the thundering implications.
I liken this to when well-known Operating Systems decide to add new features that third-party providers or vendors previously provided such as virus detection and the like. Some of those providers got wiped out. There is little need further for their wares. Others manage to stay in the game. Many do a pivot or try to stay ahead of what the big whale is doing. Etc.
Another viewpoint is that the advent of everyday users being able to customize or tailor generative AI is a good sign that maybe we will democratize AI, see my coverage on that heated topic at the link here. The notion is that the people overall will be able to direct and extend where AI is heading. That seems desirable, though a counterargument is that if those people are doing so entirely based on one particular AI offering, it doesn’t seem like as much of a democratization as ideally conceived of.
Those are deep thoughts and we can for now consider something more immediate.
The ability to make customized or tailored generative AI has the potential for riches in it, whether you do so in this case, or for what other AI makers are planning to provide. Convert your hours of frolicking on generative AI into money-making. You’ll be happy you did and your wallet will thank you for your erstwhile endeavors. Unlike seeking a gold rush, you don’t need to buy shovels and pickaxes.
You just need to put your mind to the task and go online.
Good luck with the mindful heavy lifting.
{Categories} _Category: Platforms,*ALL*{/Categories}
{URL}https://www.forbes.com/sites/lanceeliot/2023/11/07/openai-chatgpt-opens-the-door-for-you-to-make-money-by-easily-devising-income-producing-gpts-based-on-what-you-know-or-what-others-want-to-know/{/URL}
{Author}Lance Eliot, Contributor{/Author}
{Image}https://imageio.forbes.com/specials-images/imageserve/6549cd508f3be4365c7099dd/0x0.jpg?format=jpg&crop=1714,1285,×145,y0,safe&height=600&width=1200&fit=bounds{/Image}
{Keywords}AI,/ai,Innovation,/innovation,AI,/ai,Business,/business,Business,standard{/Keywords}
{Source}Platforms{/Source}
{Thumb}{/Thumb}