Recent breakthroughs in Artifical Intelligence technology have been nothing more than impressive. I used to believe that creative work should be relatively shielded by AI. I believed AI would mainly be utilized in simple automation tasks; to get rid of repetitive, numbing processes. I never thought it would turn out to challenge the creative field, but here we are!
ALL OF THE TEXT IN THIS POST IS HUMAN GENERATED. I CANT BELIEVE THAT’S A THING NOW THAT I SHOULD ANNOUNCE BUT HERE IT IS.
Let’s make one thing clear, the AI utilizes massive amounts of already made data. And everything it sprouts out is a hallucinated mish-mash of existing human generated content. Sure, if you make very unique prompts you get unique results. But that is all based on existing dataset, even though it may seem out of this world. AI just generates results so well nowadays, that it’s simply just mindblowing.
How this will take off?
As with every new groundbreaking technology that changes everything, we are going to witness wild & exciting AI technologies and services emerging and gaining mass popularity. It will provide easier ways for visual expression, productivity boost in various tasks and our chatbots are actually going to get smart – Finally! Probably it has other impact as well, we may see more AI generated stock images of people for marketing purposes for instance.
Since OpenAI and their famous product ChatGPT is tightly paired with Microsoft (It’s powered by MS + MS has heavily invested in OpenAI), we are going to see interesting development on how Microsoft’s Bing search engine gets amazing new features and perhaps long waited competitive edge over Google. I also never thought I’d be excited about Bing 😀
This certainly has caused concern in Google’s ranks. Google recently revealed their equivalent Bard unsuccesfully – resulting in 100 billion dollar loss in share value due to an error during demo.
AI technology is going to be (Actually IS) another wild west with no rules, so there are going to be also downsides. This poses great risks. It is going to be utilized for lots of bad stuff too.
We are going to get:
- …more sophisticated spam
- …more convincing phishing emails
- …more advanced misinformation campaigns
- …can be used as a powerful manipulation tool by those in power / with money.
- …copyrights and IPR violation by AI.
The uncorrupted AI
For now, we are spoiled once again with a great promises with no cost. Current AI’s are still relatively uncorrupted, they are massive and impressive free demos for the public. They are reeling us in and seducing us at this point. So enjoy while this still lasts!
But of course there’s going to be a cost in some way or another. There will be a huge commercial pressure to make profit with AI. This turning point should make us question the truthfulness of AI even more – the AI is going to be harnessed to bring profits to companies. Just because the potential is there and it’s just too tempting not to be utilized. There will still be good AI’s. But as with everything, it’s smart to be reserved and also be prepared not only for the good but also for the bad and ugly.
ART CREATED BY DALL-E2 / Prompt: “Make a 3d render of three robots dressed up as cowboys. One of them being ugly, second one of them being good and third one being bad.”
PROMPT: Make a 3d render of three robots dressed up as cowboys. One of them being ugly, second one of them being good and third one being bad.
AI-generated art
This conveniently gets us to art. There’s always been plenty of gate-keeping regarding what could be considered “art” or not. AI generated art is no different – it has sparked similar outcries as we’ve faced before. Art is a highly subjective and conveys different meanings to the eye of the beholder. I believe that AI generated art is impressive and is art by all means. A well made prompt and a good execution by the machine creates content that certainly could be considered art. No doubt about that. It may lack the human touch, but the results are still so good that machine-made may be challenging to differentiate by even artists themselves and win an art competition.
I however believe that the AI is taking unfair advantage of artists. I’m concerned about the situation where artists that have spent years and years to perfect their own unique style, may have their content essentially stolen and included in to the dataset of AI without their permission. After this the AI can generate and utilize that unique style in any prompt, and it can do it quite well. In fact, it does it so well that there exists this webpage called haveibeentrained.com where artists can check if their content has been fed to the AI or not. The AI doesn’t only steal one image by an artist, it’s able to replicate the whole style of the artist. Sounds a bit unfair to me, doesn’t it?
This has sparked fury and anger among artists and I don’t blame them. There should be rules, but as I mentioned before – it’s a wild west right now…
Can anything be trusted?
As more and more of online content can be generated by AI, we are going to face more and more fundamental trust issues in virtual world. We’ve had plenty of trust issues online since the birth of internet. But we are now facing a time where the automated content generation is going to reach industrial levels. This certainly will bring challenges. Trust will become even more important and a key factor in the future. What is the virtual world we feel content with, and is it going to be worth succumbing into?
If we travel to the future and imagine a world full of AI generated virtual content, is human generated content going to be more valuable and maybe secured as NFTs? Is blockchain going to provide some necessary means of verification for unique, human generated content? It may prove to be useful then, necessary even.
Do I feel threatened by AI?
Not yet. I see it as useful piece of tech that was bound to happen at some point. Certainly surprised by how good it is. There are however many unique, complex and detailed level problems that don’t have that existing input for current AI. ChatGPT manages to give only broad, general answers. ChatGPT also has plenty of safeguards to avoid responsibility of users actions.
I’d imagine current ChatGPT may be a useful tool or aid for searching for some common good practises for problems. This of course requires you to identify a problem that is clear and easy to formulate into a question. The key is to be able to ask the right questions. No amount of AI is helpful if the question is poorly formulated.
We all knew how to Google, now we need to learn new ways of harnessing the digital brain for our aid!