Honestly it doesn’t sound like a bad idea. However there are some concerns. AI is obviously a concept and technology that is inevitable. But both blockchain and AI are both fairly new and should be considered to remain exclusive. At least until J.U.P. has had enough development.
I don’t think it’s a good idea “now” but maybe in the future??
AI is going to be a very real and effective part of DeFi moving forward. I look forward to seeing AI be utilized and available to the masses. This is a great concept! But, as I’ve seen someone ask above, how is this project going to create revenue? Will this be a subscription service? Does the AI algorithm take a cut of the transaction?
You basically just detailed a basic approach on model training. Is there anything innovative to your approach that will really separate ZeroAI to other inference models? What is UNIQUE about ZeroAI?
As far as i know, using ANN to forecast the stock market was dated back in around 1998. I am curious about how you train the data and how many since LSTM could only learn 1000 sequences of data at best. Dated back during year 2000, I learned that no matter how good your AI model, you MUST also be very good in the market (crypto or stock) using the traditional methods.
Such a good add-on man!
Sounds amazing, but I see one big problem. Lets assume that you will reward users with zeroai tokens when your predictions are wrong. What in case black swan event or any error, bug etc where the predictions price is very different from actual price? Will you have a specific pool of tokens to give away to zeroai users, if so, how many and how you calculate this specific amount?
I just hope i am not late. I like the project. When is it launching?
It is a very interesting concept, but also sounds like it could be perceived a little dangerous. I mean I am a fan of AI and use it a lot, especially in experimenting and utility based applications, I even have an own front-loader with core code to elevate cloud based languages models. But really seeing the amount and capacity of hallucination or incremental fails, I am left thinking that whenever the AI model does correct predictions, users will likely attribute it as a given base standard, while as soon as only a little goes wrong they will get the pitchforks. A bit about the 80-20 rule mindset. Or the fact that negative feedback is always ten times higher than positive.
I love this. People’s index funds are already dominated by algos. I would love to be able to give zeroai the ability to execute trades within certain parameters and risk tolerances. This should probably (for safety) be very conservative. I really see this being incredibly disruptive to passive investing strategies that have completely grown to dominate tradfi and leaves an enormous amount of alpha on the table. This could be the vanguard of web3 just as easily as easily tools for traders. Id love to see how training goes and allow accuracy to be tracked publicly. Eventually, this will hit a point to be in charge of pensions.
Information asymmetry is the largest handicap that tradfi has had over the general public and has used it to fleece retail investors and transform everyone into their own hedge fund managers. This can be the most radical and empowering project ive seen on lfg. The question becomes not whether or not the project is good, but whether or not you guys can deliver. Fantastic project, and if youre able to share more on where you are now, your roadmap, and your technical qualifications as well as your current successes to date and identified obstacles, all the better. I am routing for this one.
I’d also like to hear what advantages you think you have over other projects trying to develop similar strategies and if you think jupiter or solana ecosystem makes you more competitive. Maybe allow jupiter to handle trade execution?
Also, does the best ai dominate the others? Will the project thats best funded take the lead and front run all the available alpha? Theres plenty of humans to outperform for a while, but just wondering.
Lastly, just a suggestion that allowing zeroai to execute trades for a .02% fee (half the cost of s&p etf indexes) could help fund continued development. The hardest thing for normies getting into web3 is how much involvement it requires. Tradfi set and forget is just too appealing. The token could also allow for trading that is taxed advantaged. So you only claim a gain when you sell the token while zero ai trades in the background. Sorry… i understand this isnt the project you proposed but excited by the possibilities of the tech.
I’m pleased to see a platform focused on AI being presented here.
I look forward to doing my research on this. Looks like a cool and useful idea. Fingers crossed
ATB😁
In my experience, managing user expectations and transparently communicating the limitations and capabilities of AI are crucial. As developers and proponents of AI, we must emphasize that while AI can significantly augment our capabilities, it is not infallible. Continuous learning, user education, and robust feedback mechanisms are essential to navigate these challenges and foster a balanced understanding of AI’s role and potential.
Absolutely! Drawing from my own knowledge and experience, I wholeheartedly agree. Managing user expectations and transparently communicating AI’s capabilities and limitations are pivotal. While AI can indeed enhance our abilities, it’s essential to recognize that it’s not infallible. Continuous learning, educating users, and actively seeking feedback contribute to a more balanced understanding of AI’s potential and its place in our lives.
Hello,
Sounds interresting!
If in the future, your tech is integrated to DEX or agregators like Jupiter, and if lots of people use it, don’t you think that there will be an effect on the prices? Couldn’t it amplify panic buy/sell movements?
Then how could your model take this in consideration?