Learn: Crypto and AI- A yay or nay combination?
A contrast between AI and crypto is that AI requires trust, while crypto emphasizes verification. Projects working to combine AI and crypto demonstrate that the two sectors are compatible.
The breakout tech story of the moment is around ChatGPT, which has been garnering significant interest and a lot of positive commentaries.
Developed by the research laboratory, OpenAI, ChatGPT operates like a very advanced AI chatbot. Having been trained on huge samples of content, it can interact with people in natural human language, and provide detailed responses to complex queries.
ChatGPT can be utilized in lieu of a search engine, providing a direct answer to questions, rather than links to websites, but it can also be instructed to create content in specified styles.
Applications of the latter utility include the writing of computer code, academic essays, articles and reports, tweet threads, and even some attempts at poetry.
There’s a strong contrast between sentiment around ChatGPT, compared with everything that has been happening in crypto.
While crypto news has, for several months, revolved around catastrophe and collapse, with the likes of FTX dragging reputations through the mud, ChatGPT is being greeted with optimistic enthusiasm, and provides a breath of fresh air from the doom and gloom of the crypto markets.
Looking further at both crypto and AI, other contrasts between the two tech sectors become apparent, along with possibilities that they might draw together and intertwine.
A key difference between AI and crypto, is, broadly speaking, that AI represents trust over verification, while crypto emphasizes verification over trust.
When you ask a question to ChatGPT, in the way that you might to a search engine, what you get in response is an authoritatively-phrased, human-sounding answer.
What you don’t get is primary sources, or, in fact, any citations at all. You simply have to take the answer at face value, trusting that whatever goes on under the AI hood lands on an accurate response.
And, as it goes, ChatGPT’s answers can sometimes still, at this stage, sound plausible and confident but actually be inaccurate or inconsistent (ask the same question multiple times and you might get multiple different answers).
As AI improves, it should become more reliable, but still, there is the black box element: even the people who create AI may not always know precisely what is going on inside the system on the way to the provision of an output.
And, there is the issue of garbage in, garbage out, which is a concern around all kinds of computer modeling. Output quality depends entirely on the quality of the data going in, and AI cannot really think about anything. It takes in data and is trained to model a consistent replication or summary of what it has absorbed.
A caveat here is that this criticism is primarily considering ChatGPT as a way of finding information. However, as a means to create preliminary outlines for texts and code, it is more immediately useful, AI imagery could quickly replace some stock photography and simple graphic design, and there is a wealth of other AI research happening around the world.
However, the point is that AI, in general, requires trust in its method and that we are willing to believe in the veracity of its output.
And, if you observe fictional representations of very advanced AI (which are worth keeping in mind, since they depict thinkable paths), what you tend to find is computers providing humans with nimble but unverified assistance.
Crypto, by contrast, is the antithesis of this approach, as it places at its core the idea that we should trust nothing at all. Rather, the open and distributed blockchain process will transparently verify and lock-in transactions, and is hard-coded to have no capacity for manipulation or under-the-hood mystery.
There are no secrets about what occurs, and no mutability, since Bitcoin (the original cryptocurrency from which all others followed) was conceived as a direct response to corruption within banking, finance and money itself.
Bitcoin hoped to replace these flawed systems with an alternative that was trustworthy precisely because it rejected the very notion of trust.
As such, crypto purports to move us from 'don’t be evil', to 'can’t be evil', while AI technology seems, at times, more like it’s heading towards, I hope nothing evil is going on in there.
It should be added, by the way, that this summary refers to crypto itself, the working blockchain technologies, and not to centralized trading platforms that have subsequently emerged, collapsed and caused immense damage.
There are projects where crypto and AI overlap, which have potential due, in part, to the contrast between the two fields. After all, imagine being able to combine AI’s flexibility and wide vista of application, with the decentralization and digital ownership enabled by crypto.
Blockchain technology has, at times, been criticized for a perceived lack of utility, but it can bolster AI, facilitating monetization and accessibility through decentralized AI platforms and marketplaces, and enabling personal ownership of trainable, tradable AI services, bots and other creations.
Some projects that are working to tie together crypto and AI include SingularityNET, DeepBrain Chain, Fetch.ai, and, in the NFT sector, Altered State Machine and Alethea.
One final contrast between the two sectors, crypto and AI, lies in the ways that they are currently perceived. Crypto has spent much of 2022 mired in scandal, and recently, its most untrustworthy actors have occupied the spotlight.
AI, on the other hand, comes across as purely tech-driven. Crypto would benefit greatly from emulating some of these characteristics, as it works to leave recklessness behind, and instead, put demonstrable utility upfront and center stage.