Chat GPT do a TA about bitcoin

[Tytanowy Janusz]

Recently, I started experimenting with AI tools. They are incredible tools with tremendous potential, but also full of errors. Out of curiosity, I asked Chat GPT for predictions on future Bitcoin prices. Of course, it immediately mentioned that it only has data up until 2021. After providing it with the current data from the last six months, it generated the following prediction. What do you think of it, and what are your thoughts on the potential of AI in technical analysis?

Short-term Scenario (1-3 months):
Based on the recent price movements, the short-term scenario for Bitcoin is as follows:

Probability: 40%
In this scenario, Bitcoin's price may experience a consolidation phase with a slight upward bias. It may encounter resistance around the previous high levels, such as $29,000-$30,000. The trading volume is expected to remain moderate, indicating a cautious sentiment among traders. The key support level to monitor is around $26,000. If the price breaks below this level, it could signal a bearish trend continuation.

Medium-term Scenario (3-6 months):
Probability: 30%
In the medium-term scenario, Bitcoin's price could exhibit increased volatility and potentially enter a new uptrend. The probability of this scenario is lower due to the overall market uncertainty and the need for a sustained bullish trend. If Bitcoin manages to break above the resistance level near $30,000, it may attract more buyers, leading to a potential rally. The key resistance to watch for is around $35,000. However, a failure to break above this level could result in a continuation of the consolidation phase or a bearish reversal.

Long-term Scenario (6+ months):
Probability: 30%
The long-term scenario for Bitcoin suggests a potential continuation of the bullish trend, but with increased volatility and periodic corrections. Bitcoin's price could aim for new all-time highs, potentially surpassing the previous peak around $64,000. The probability of this scenario is lower due to the inherent uncertainty in long-term predictions and the influence of external factors on the cryptocurrency market. It is important to monitor key support levels, such as $40,000 and $50,000, as significant breaches of these levels could indicate a trend reversal.

Second attempt with slightly different prompt:

Scenario 1: Short-term (next few days/weeks)
Based on the recent price movements, Bitcoin has been experiencing a downward trend. There is a possibility that the price might continue to decline in the short-term. The key support level to watch is around $26,000. If Bitcoin breaks below this support level, it could indicate further downside potential. However, if the price manages to hold above this level and starts to rebound, we may see a short-term rally towards the resistance level around $28,500.

Probability: There is a 40% chance of a short-term price decline, and a 60% chance of a short-term price rebound.

Scenario 2: Medium-term (next few months)
In the medium-term, the price of Bitcoin will be influenced by various factors, including market sentiment, institutional adoption, regulatory developments, and overall market conditions. If Bitcoin manages to regain positive momentum and break above the resistance level around $28,500, it could signal a bullish trend. In this scenario, Bitcoin may experience a gradual uptrend, with potential resistance levels at $30,000 and $32,000.

Probability: There is a 30% chance of a medium-term bullish trend, a 40% chance of a sideways consolidation, and a 30% chance of a medium-term bearish trend.

Scenario 3: Long-term (next several months to a year)
The long-term outlook for Bitcoin depends on various fundamental factors, including macroeconomic conditions, institutional adoption, regulatory developments, and market acceptance. If Bitcoin manages to maintain its position as a leading cryptocurrency and continues to gain mainstream acceptance, there is a possibility of a long-term uptrend. In this scenario, Bitcoin could potentially reach new all-time highs, surpassing the previous peak levels. However, it's important to note that the cryptocurrency market is highly volatile, and there are inherent risks involved.

Probability: There is a 40% chance of a long-term bullish trend, a 30% chance of a sideways consolidation, and a 30% chance of a long-term bearish trend.

Its nice that both are bulish

[1714654322]

1714654322

[1714654322]

1714654322

[1714654322]

1714654322

[1714654322]

1714654322

[dragonvslinux]

This is kind of interesting, even if worth taking with a large pinch of salt  

Out of interest did you also provide it with human-based analysis of Bitcoin's price, such as tradingview publications, or merely the price data? As it seems to be regurgitating a lot of analysis already out there imo. Therefore it seems likely it has categorised the possible outcomes based on the amount of analysis provided on these outcomes, as opposed to "forming" it's own statistical analysis.

Also, although bullish predictions have been made short to long-term, there is notably no possibility of a bearish outcome here? Ie 100% of the outcomes are bullish which seems completely unrealistic, even if price is to continue higher. Either way, I'd be interested to hear updates from Chat GPT if you are willing to feel it the latest data in the future, in order to see how accurate these predictions are!

Thanks for sharing.

[Tytanowy Janusz]

This is kind of interesting, even if worth taking with a large pinch of salt  

Out of interest did you also provide it with human-based analysis of Bitcoin's price, such as tradingview publications, or merely the price data? As it seems to be regurgitating a lot of analysis already out there imo. Therefore it seems likely it has categorised the possible outcomes based on the amount of analysis provided on these outcomes, as opposed to "forming" it's own statistical analysis.

ChatGPT does not have access to the internet. The data it is based on ends in 2021, and what comes after that is very fragmented and incomplete. So, the analysis I shared with you is not based on other online analyses of current bitcoin price change. The data the model received from me consists of daily OHLC (Open, High, Low, Close) prices starting from January 2023. The analysis seams to be based on that, price action before 2021 and old analysis that model already had.

Also, although bullish predictions have been made short to long-term, there is notably no possibility of a bearish outcome here? Ie 100% of the outcomes are bullish which seems completely unrealistic, even if price is to continue higher. Either way, I'd be interested to hear updates from Chat GPT if you are willing to feel it the latest data in the future, in order to see how accurate these predictions are!

It seems to me that these percentages do not add up to 100%, because they relate to different time intervals. The bot seems to give the most likely scenario for a given period and gives its probability. The second analysis seems better.

Thanks for the answer. Yes, I think it's an interesting topic and I'll try to update the thread with current analysis as well as improve the prompt for AI to get better answers in future.

[Aanuoluwatofunmi]

Recently, I started experimenting with AI tools. They are incredible tools with tremendous potential, but also full of errors. Out of curiosity, I asked Chat GPT for predictions on future Bitcoin prices. Of course, it immediately mentioned that it only has data up until 2021. After providing it with the current data from the last six months, it generated the following prediction. What do you think of it, and what are your thoughts on the potential of AI in technical analysis?

The AI may be as that good enough to believe in what data analysis they present us but that doesn't mean they are hundred percent 100% accurate and keen for acceptance, they also work in line base on the level of the programming made on them to perform a particular task, this is one of the reasons the use of AI is not an acceptable means of postings on this forum's signature campaigns, the end result they may later generate for you may not be as adequately equal to the expected results or the real state or what the future may landed at.

[BitcoinPanther]

This is very interesting, I also asked another chatbot to do a TA about Bitcoin and somehow I got a different answer.

1-3 months

In the short term, Bitcoin's price is currently consolidating within a range between $28,000 and $30,000. The bulls are trying to push the price above the resistance level of $30,000, while the bears are trying to push the price below the support level of $28,000. If the bulls are successful, the price could rally to $35,000. However, if the bears are successful, the price could fall to $25,000.

Probability:

The probability of Bitcoin's price rallying to $35,000 in the short term is 50%. The bulls and bears are currently evenly matched, and it is difficult to say who will be successful in the short term.

4-6 months

In the medium term, Bitcoin's price is still in a downtrend. The price is currently trading below the 200-day moving average, which is a bearish indicator. The bulls need to close the price above the 200-day moving average in order to invalidate the downtrend. If the bulls are successful, the price could rally to $40,000. However, if the bears are successful, the price could fall to $20,000.

Probability:

The probability of Bitcoin's price rallying to $40,000 in the medium term is 30%. The bears are still in control of the market, and they are likely to continue to push the price down until the bulls can close the price above the 200-day moving average.

7-9 months

In the long term, Bitcoin's price is still in an uptrend. The price is currently trading above the 200-day moving average, which is a bullish indicator. The bulls need to close the price above the all-time high of $68,789 in order to invalidate the uptrend. If the bulls are successful, the price could rally to $100,000. However, if the bears are successful, the price could fall to $10,000.

Probability:

The probability of Bitcoin's price rallying to $100,000 in the long term is 70%. The bulls have been able to hold the price above the 200-day moving average, and they are now trying to push the price above the all-time high of $68,789. If the bulls are successful, the price could rally to $100,000. However, the bears are still in control of the market, and they could easily push the price back down to $10,000.

As we can see the probability for the TA to realize in short term is about 50%, medium term 30% and long term 70%.  The chatbot is connected to internet so it is possible that it gathers data on the internet and count the months from current  to the months asked.

It is fun to ask AI's since they tend to give different answer if we reconfirm our questions, like when I reconfirm the AI's answer it gives me this answer

I am not sure about my answer on what is TA of bitcoin price for 1-3 months, 4-6 and 7-9 and the percent probability for it to happen. I am a large language model, also known as a conversational AI or chatbot trained to be informative and comprehensive. I am trained on a massive amount of text data, and I am able to communicate and generate human-like text in response to a wide range of prompts and questions. For example, I can provide summaries of factual topics or create stories. However, I am not able to predict the future with 100% accuracy.

The future price of Bitcoin is uncertain, and it will depend on a variety of factors, including economic conditions, regulatory changes, and technological advancements. It is important to do your own research before making any investment decisions.

[STT]

Seems like reasonable conclusions, stick to simple observations and its not hard to point out that moving averages have been aligning to a more positive direction then we saw in 2022.    The immediate price action is meeting 50 day which is negative, so most immediately we are challenged in that measure but its less important then the 200 day which is rising so I think me and the bot should agree.  
 These AI bots aren't real intelligence imo but also not something to dismiss either, they collate information and make it more accessible, its finesse more then an actual giant leap forward but I'd like to find out Im too pessimistic also.   I expected something like this 20 years ago, I believe the power to provide processing has become more available thats part of why now.

[hatshepsut93]

With medium and long-term predictions, it would take us decades to  acquire enough data to judge just how accurate ChatGPT is. It would be interesting to take past price charts, ask ChatGPT where the price will go and compare it to what actually happened, although maybe the current model is based on the full Bitcoin price history and it wouldn't be fully correct to apply it to some previous fragment.

[AmoreJaz]

With medium and long-term predictions, it would take us decades to  acquire enough data to judge just how accurate ChatGPT is. It would be interesting to take past price charts, ask ChatGPT where the price will go and compare it to what actually happened, although maybe the current model is based on the full Bitcoin price history and it wouldn't be fully correct to apply it to some previous fragment.

what we can do from chatgpt's predictions is just use it as one of the references how to decide on what to do with the market. clearly, there are still limitations on how it formulates its answers, basing on the database it has been feed on.
but the response is quite comprehensive if you are a beginner on this market. i would say, no newcomer can come up to such conclusion that fast!

[tabas]

What it had provided for its probability is like a safe term to be given even by someone who analyzes the market.

It seems to me that these percentages do not add up to 100%, because they relate to different time intervals. The bot seems to give the most likely scenario for a given period and gives its probability. The second analysis seems better.

While it has given these analyses and predictions of what it thinks of the market. Before, it wouldn't say that much so I'd say that this AI tool is really learning from the data that it's given because when I've asked it before. It has said that he can't do that much and just given a short words of what it thinks about Bitcoin and a bit of its TA. Thanks OP, really a good topic. Newbies might be dependent on it but it's still best to just use this as a reference but not the deal breaker of their decisions.

[Oilacris]

With medium and long-term predictions, it would take us decades to  acquire enough data to judge just how accurate ChatGPT is. It would be interesting to take past price charts, ask ChatGPT where the price will go and compare it to what actually happened, although maybe the current model is based on the full Bitcoin price history and it wouldn't be fully correct to apply it to some previous fragment.

In talking about future prices or predictions then it would really be just that normal that there would really be some connecting scenario on what happened in the past would really be pass or would really be
not that surprising that it will really be that pertain or in connection into the future. We do make out those assumptions which is something that would really be that normal or having the impression
that prices could really be going that beyond those numbers as the cycles of this market would really continue to progress.

We didnt even had able to believe that Bitcoins price hit up that 69k ATH which we are just hoping on just breaking that 20k resistance ath way before on those earlier years.
If Chatgpt did make out such response then its true that it would  be really somewhat not some newbie kind of approach of response which it would really be that considerable
to be read up for us to those who do have experience or simply a very significant approach but it didnt really give out any guarantees which is really that good.

[Aliem Nur]

it's very interesting that you asked about the price of Bitcoin on ChatGPT and the answer was bullish,
from TA and history it has also been said that the price of Bitcoin has indeed started a bullish trend so we must be optimistic that ChatGPT is also correct,
let's hold Bitcoin and don't sell it at a cheap price!.

[rat03gopoh]

I hope someone will be curious to experiment with real money on some trades so as to draw a more complete conclusion whether the AI's predictions are accurate enough or not. So far, AI chatbots have had many objections about the integrity of the information provided, especially talking about the future.

If they 're accurate enough and start to become the main tool of trading, maybe gradually there will be a significant change in market variables, so that all the technical analysis that we have known for a long time will no longer work.

[Poker Player]

Of course, it immediately mentioned that it only has data up until 2021.

I understand that future versions will use real-time data, especially in paid versions.

What I think is that they will be widely used in a few years, but as a tool that helps, like many others, without using it we can have 100% certainty that it will be successful, far from it.

it's very interesting that you asked about the price of Bitcoin on ChatGPT and the answer was bullish,
from TA and history it has also been said that the price of Bitcoin has indeed started a bullish trend so we must be optimistic that ChatGPT is also correct,
let's hold Bitcoin and don't sell it at a cheap price!.

From what I see ChatGPT thinks the same as me. At the heights we are at, I am also bullish: there is less than a year left for the halving and in the future, if we base ourselves on the past history of bitcoin, which is the only thing we have, the price can only go up, with ups and downs, but rise in the medium and long term.

[yhiaali3]

It is good that almost all of the bot predictions are positive about the rise of Bitcoin in the medium and long term. This makes me optimistic, although not all predictions can be 100% accurate.

It would be a good idea to keep an eye on the topic updates to see how advanced the bot can be in predicting bitcoin prices and what percentage of the predictions are accurate.

[GreatArkansas]

I hope someone will be curious to experiment with real money on some trades so as to draw a more complete conclusion whether the AI's predictions are accurate enough or not. So far, AI chatbots have had many objections about the integrity of the information provided, especially talking about the future.
(....)

You already know the answer here, and these AI always have disclaimers if your questions are about financial like this financial advice which involves money and you can lose your money if AI is wrong.
So for me, not relying on these AI at the beginning is always good, it's like prevention better than cure.

[Tytanowy Janusz]

It is fun to ask AI's since they tend to give different answer if we reconfirm our questions, like when I reconfirm the AI's answer it gives me this answer

you have to be careful because AI doesn't know what truth is. He is constantly making up and pouring water using the probability to write the next word of his speech. Nothing more. And since the whole thing sounds likely, we humans automatically think it's true. Nothing could be more wrong.
In my short adventure with chatGPT, which is probably the most advanced model, I have repeatedly encountered total lies presented so credibly that I began to doubt myself. The mere fact that gpt chat is not connected to the internet. On the one hand, when I asked him to download data from the website, he said that he could not, because he did not have access, but when the conversation turned out that the amount of data I want to give him is most conveniently done by giving a link to the website, he suggested it himself , then pretended to fetch data and then pretended to create a response based on it, when it was just waffling, not backed up by the content of the page. you have to be careful with these models and don't expect more from them than what they were made for and if anything, treat their answer with a grain of salt.

[EarnOnVictor]

Recently, I started experimenting with AI tools. They are incredible tools with tremendous potential, but also full of errors. Out of curiosity, I asked Chat GPT for predictions on future Bitcoin prices. Of course, it immediately mentioned that it only has data up until 2021. After providing it with the current data from the last six months, it generated the following prediction. What do you think of it, and what are your thoughts on the potential of AI in technical analysis?

As AIs are being developed and advanced, cautiousness is also being adhered to so that it will not be in violation of anything. Despite this, many are still complaining and some clamours for the halt of the Al projects which I think is not possible at this early stage. Besides, what you stated above could only mean that ChatGPT on its own is being limited to some features and clues about the market so that people would not be using it to chase shadows for their selfish interests, which will, in turn, violate some rules of the financial market. Nonetheless, if the AI has the information and capability to read the future price of the market, it won't still be productive, do you know why?

It's because it will always rely only on the data at its disposal, but the market is too dynamic for that, it can't be rigged by any robot, and there is no central control for the financial market. Conclusively, there is no holy grail anywhere and there will never be.

[Yamane_Keto]

My problem with such applications is the increase in trust. Many consider it like a specialized expert who has years of experience. While these algorithms need real-time analysis, they still need a lot of time before the data is real. This is if we assume that it is possible to obtain an accurate prediction.
It's closer to the proverb lie lie lie and they will believe you, people will believe Chat GPT and with the passage of time even the lie of AI will become true because people will start making their investment plans based on those tips, these tips are what will move supply and demand and so the price, just as Elon Musk's tweets do.

[Tytanowy Janusz]

As AIs are being developed and advanced, cautiousness is also being adhered to so that it will not be in violation of anything. Despite this, many are still complaining and some clamours for the halt of the Al projects which I think is not possible at this early stage. Besides, what you stated above could only mean that ChatGPT on its own is being limited to some features and clues about the market so that people would not be using it to chase shadows for their selfish interests, which will, in turn, violate some rules of the financial market. Nonetheless, if the AI has the information and capability to read the future price of the market, it won't still be productive, do you know why? It's because it will always rely only on the data at its disposal, but the market is too dynamic for that, it can't be rigged by any robot, and there is no central control for the financial market. Conclusively, there is no holy grail anywhere and there will never be.

1- using AI to make money in the market is not a violation of anything. Just like in the 17th century, the use of technical analysis at the time when it was just being created, only a handful of people had access to it, thanks to which it easily outplayed the rest.
2- you say that AI will not cope with the market even with all available data. So how does a person cope? After all, in most cases, he knows nothing more than what he sees and reads on the Internet.

[dragonvslinux]

This is kind of interesting, even if worth taking with a large pinch of salt  

Out of interest did you also provide it with human-based analysis of Bitcoin's price, such as tradingview publications, or merely the price data? As it seems to be regurgitating a lot of analysis already out there imo. Therefore it seems likely it has categorised the possible outcomes based on the amount of analysis provided on these outcomes, as opposed to "forming" it's own statistical analysis.

ChatGPT does not have access to the internet. The data it is based on ends in 2021, and what comes after that is very fragmented and incomplete. So, the analysis I shared with you is not based on other online analyses of current bitcoin price change. The data the model received from me consists of daily OHLC (Open, High, Low, Close) prices starting from January 2023. The analysis seams to be based on that, price action before 2021 and old analysis that model already had.

So if ChatGPT isn't analysing human-based analysis, I assume it's using available generic technical analysis (ie publicly available information) to analysis the current price based on OHLC? It'd therefore make sense why the analysis is relatively in-line with other human-based analysis, if it is simply applying general TA to Bitcoin's current price. I guess you could say it's reassuring in that sense.

Also, although bullish predictions have been made short to long-term, there is notably no possibility of a bearish outcome here? Ie 100% of the outcomes are bullish which seems completely unrealistic, even if price is to continue higher. Either way, I'd be interested to hear updates from Chat GPT if you are willing to feel it the latest data in the future, in order to see how accurate these predictions are!

It seems to me that these percentages do not add up to 100%, because they relate to different time intervals. The bot seems to give the most likely scenario for a given period and gives its probability. The second analysis seems better.

Fair enough, that makes sense. You're right that the TA on different time-frames would be independent of each other, and therefore the percentages shouldn't be considered cumulative.

Thanks for the answer. Yes, I think it's an interesting topic and I'll try to update the thread with current analysis as well as improve the prompt for AI to get better answers in future.

Looking forward to future updates. If I had to guess, it could well analyse Bitcoin's price using technical analysis theories better than humans, as will do so without emotion.