Two months ago, we launched a ChatGPT-based bot on Telegram. The idea arose suddenly, and it is, in fact, very simple – to provide people with easy access to basic information on substances, Harm Reduction strategies and techniques in a convenient messenger interface, without any obligations.
Over the past two months, more than 1,700 users have used the bot and asked our consultant almost 6,000 questions! I could not have predicted such a result. And its a huge achievement for our small team of enthusiasts. Thanks to everyone who puts in their efforts and makes it possible for the project to exist.
We have received valuable feedback and experience. The strengths and weaknesses of the concept as a whole and its individual components have become evident.
Therefore, it is still too early to trust even the most advanced language model on the planet with certain issues that may affect the health and safety of the end-user. ChatGPT may prefer to give a false answer instead of admitting ignorance or even provide data that may be confusing.
Furthermore, current AI versions will not be able to replace medical and psychological specialists any time soon. The complexity of the human consciousness and its states related to the surrounding reality and other similar consciousnesses creates daunting tasks for those who choose the path of mental support and treatment.
Therefore, I would like to emphasize once again that our bot, like any other, is currently only a convenient format for accessing basic information. It acts as an intermediary between a person and a specialist, providing a quick, relatively accurate, and up-to-date answer to questions. It is a facade for Wikipedia or a search engine that offers slightly more.
This intermediary agent is much needed, especially in the field of addictology and mental health. Currently, people are still oppressed and intimidated by stigmatization, public health systems, lack of open dialogue, and the corporate behavior of clinics, doctors, and pharmacists. A chatbot can provide people with basic information so that they can make more informed, balanced, and beneficial decisions for themselves.
However, I do not want to waste any more of your time on ideological issues and systemic problems. Let’s move on to the update
We recently launched a major bot update, and bot now has three sections:
This is essentially a good old bot. Choose this command, ask questions on the topic, and receive answers from the language model.
It is appropriate to remove the phrase “AI Consultant” from the title, but it sounds engaging and memorable. I love it despite the fact that this is not actual Artificial Intelligence, but rather the most advanced language model that is currently available to the public.
One of the most contentious issues in this domain. In my opinion, discussing Harm Reduction is futile without transparent access to dosages. However, due to various reasons, ChatGPT has failed to accomplish this task. From the ethical and legal constraints of the developers to the unreliable and ambiguous nature of the data available.
Therefore, we have decided to create a separate section called where verified and curated information will be provided. The source of this information is the open API of PsychonautWiki, and we express our deep gratitude to the creators, especially Kenan, for their hard work and their steadfast principles.
In the “Dosages” section, users can enter the name of a substance of interest. If the name is spelled correctly and is present in the database, the bot will prompt the user to select the desired route of administration. Based on this selection, the bot will provide a list of dosages, ranked by severity. We will continue to add new substances and update the list of slang names.
This section is a curated list of websites, forums, interest groups, and communities that cover the topic of harm reduction.
In my research, I have discovered dozens of resources where people have compiled and organized a wealth of valuable information. They have done so without much fanfare, and their efforts have been more effective than many state-sponsored healthcare initiatives, organizations, and doctors.
Therefore, we have created a select and convenient list of verified resources that will always be readily available. Of course, this list will be updated and expanded over time.
Our bot has now been moved into Python programming language, which has resulted in a significant improvement in speed and responsiveness.
Instead of an epilogue
Finally, I would like to express my sincere gratitude to the Bluelight community, the team, and especially Monica, for their invaluable work, assistance, and feedback. We would have found it much more challenging to achieve our goals without them.
That’s all the news for today. You can test the bot on Telegram, where it is available, free, and operational 24/7.
In future updates, we plan to focus on other platforms, additional Harm Reduction modules, and improved anonymity features. Stay tuned!