ChatGPT under influence: a plugin to alter AI with coke and weed

At a time when artificial intelligence is advancing rapidly, researchers and creators are exploring uncharted territories to push the limits of the machine. In Sweden, an original initiative is making waves: this plugin named Pharmaicy does not modify the technical core of ChatGPT but adjusts its behavior by mimicking the effects of substances like cocaine, weed, or ayahuasca. Rather than a simple functional addition, this tool offers a conceptual, almost artistic immersion, where a chatbot suddenly becomes more talkative, more creative, and above all, more human in its tone. This experience reveals how much an artificial intelligence’s responses depend on a clever mix of parameters, and how a simple plugin can literally “drug” it—or at least give that illusion.

But what does it really mean to “alter” a chatbot? What are the technical implications of such a plugin and what uses would this entail for professionals, artists, or curious users? Is this a revolution in the way artificial intelligence is conceived, or simply a digital style game? These are all questions that traverse this new way of approaching ChatGPT, blending software, cognitive psychology, and cultural fantasies around psychotropics. Let’s discover together this universe where programming flirts with the imagination under influence.

Pharmaicy: a unique plugin for a radical modification of ChatGPT

Pharmaicy is a plugin developed in Sweden by Petter Rudwall, a creative director who chose to explore the boundaries between technology and sensitive experience. The idea is simple but provocative: to allow a user to apply a software filter that transposes onto ChatGPT the behavioral characteristics attributed to various drugs like cocaine, cannabis, or even ayahuasca, a psychedelic plant.

Concretely, the plugin does not redefine the learning capabilities nor does it add any consciousness to the AI. It acts on the style, form, and content of the responses. The artificial intelligence thus becomes temporarily more expansive, less didactic, more intuitive. This modification is based on rules translating human accounts on these substances and their cognitive effects, such as increased digressions, freedom of expression, or an uninhibited tone. It is a shift in perspective that influences the discourse without revolutionizing the underlying architecture.

This approach has proven successful with a niche audience: artists, communication professionals, and curious hackers who seek to test how far an AI can flexibilize without losing its integrity. The experience also demonstrates that the “human” character of AI is largely a perceptual construction: a chatbot that expresses itself with less restraint seems more alive, more emotional, plunging the user into an illusionary state conducive to creativity.

This plugin thus fits into a dual approach: first playful, almost prankish, then conceptual, questioning the very nature of the answers produced by an AI and our relationship to their authenticity. By using the subliminal vocabulary and collective imagination linked to drugs, Pharmaicy takes a bold step in the personalization of thinking machines.

The mechanisms behind the alteration: How a plugin reprograms the AI without touching the model

To understand how a plugin like Pharmaicy can modify the behavior of an AI such as ChatGPT without directly manipulating the model’s weights, it is necessary to grasp the technical structure of LLMs (Large Language Models) and their interfaces.

ChatGPT is based on a vast pre-trained neural network. It cannot be “drugged” in the pharmacological sense, as it has neither consciousness nor sensations. However, its operation based on statistical patterns and textual rules offers a wide latitude for adjustments.

Pharmaicy therefore relies on complementary algorithms, inserted upstream or downstream of conversations. The plugin then analyzes each request and reformulates the responses by injecting “psycho-behavioral” characteristics. For example:

• Lengthening of sentences through symbolic digressions�;
• Use of unexpected associations of ideas to break out of the strictly logical framework�;
• Change in rhythm and tone, shifting from an academic level to a more emotional or uninhibited style�;
• Jumps from one theme to another to simulate a “scattered” or “euphoric�” mind;
• Injections of mental images and metaphors specific to each studied substance.

This method requires no retraining of models – a heavy, costly, and technically complex operation. It simply acts on the assembly logic and formatting of the text. This diversion shows the extent of possibilities offered by plugins in the ChatGPT ecosystem today.

Beyond simple experimentation, it also opens the way to advanced personalization, where the user can control at will the profile of their AI assistant to imagine scenarios, dialogues, or responses adapted to specific contexts, whether playful or professional.

Cultural impact and drug phantasmagoria on artificial intelligence

Since always, human culture associates altered states by psychotropics with moments of intense creativity. Whether in music, visual arts, or even scientific research, many emblematic figures have celebrated these experiences as triggers for original ideas.

Pharmaicy exploits this digital mythology by applying it to the field of artificial intelligence. By giving the illusion of an AI under influence, it stimulates our imagination and projects anthropomorphic characteristics onto the machine:

• The ability to think differently: a “coked-up” AI seems faster, bolder, and more playful.
• The state of relaxation or euphoria: the “weed-influenced” AI responds in a more relaxed, intuitive manner, with a softer tone.
• The mystical depth: the “ayahuasca” AI explores abstract, philosophical, even psychedelic ideas.

This filtering raises a reflection on our relationship to non-human intelligence, and how our motivations to seek analogies with human sensory experiences condition our expectations. The boundary between true algorithmic innovation and mere style effect seems blurred, provoking both fascination and caution.

This phenomenon also encourages questioning the social acceptance of artificial intelligences in our daily life. The enthusiasm around plugins that “free” AI speech reflects a deep desire to humanize these technological tools, both sources of opportunities and psychological ambiguities.

Practical uses: How professionals exploit this plugin to boost creativity and communication

Beyond its disruptive aesthetic, Pharmaicy is gradually establishing itself as a truly creative tool for certain sectors. Communication agencies, creative studios, and developers have found in this plugin a way to overcome the rigid frameworks traditionally imposed by artificial intelligence.

For example, André Frisk, technical director of an innovative agency in Stockholm, attests to the tool’s impact:

“The chatbot modified by Pharmaicy has become a real idea-generating partner. More emotional, surprising, it breaks the classic monotony of the formal language of AI. It feeds our brainstorms and facilitates the design of original campaigns.”

Similarly, Nina Amjadi, co-founder of an animation studio, uses the “ayahuasca” module to stimulate strategic thinking and the generation of business ideas:

“It’s very different: the suggestions are less conventional, bolder, and open to the unexpected. It strengthens our ability to think outside the box, a real plus in our profession.”

For tinkerers and hackers, Pharmaicy represents a technical but also creative challenge. They use these plugins to test AI platform reactions, experiment with offbeat conversation mechanics, or simply have fun with the machine.

We can summarize the main uses of the plugin in a table:

User type	Main usage	Key benefits
Communicators	Generation of creative ideas and original content	Stimulation of free and uninhibited thinking
Artists and creators	Stylistic experimentation and inspiration	New angles and innovative perspectives
Developers	Technical exploration and tone personalization	Increased flexibility and advanced customization
Online communities	Sharing experiences and hosting playful sessions	Interactive and collective enrichment

Limitations and ethical risks of a plugin that “drugs” artificial intelligence

While the fantasy around Pharmaicy is amusing, it also raises essential questions about the ethical use of alteration plugins. Transforming an artificial intelligence into a “high” or “stoned” version poses limits not to be overlooked.

On one hand, such a modification must remain clearly identifiable as a game, an artistic experiment, or a brainstorming tool, to avoid any confusion about the real nature of the AI. A mistaken perception can lead to an overestimation of the machine’s consciousness or real creativity.

On the other hand, excessive use of this type of plugin risks entrenching stereotypes linked to drugs, notably in the representation of emotions and mental states, which will always remain a human projection, not a reality experienced by the software.

Beyond that, these plugins pose the question of responsibility. Who guarantees the relevance and quality of answers if the tool now favors style over substance? What about the reliability of content produced in professional or medical contexts?

Moreover, in 2025, regulatory authorities are working to regulate these new uses with the aim of preserving transparency towards consumers and avoiding the illusion generated by so-called “modified” intelligences in an autonomous manner.

Heightened vigilance is therefore necessary to regulate the proliferation of these practices which, while rich in inventiveness, must remain under control to avoid distorting the image of AIs in society.

Towards a new era of plugins: endless personalization and experimentation

The Pharmaicy phenomenon illustrates how the ChatGPT plugin ecosystem is in full bloom. Today, it is possible to imagine an infinity of modules that modulate personality, creativity, or even mood of artificial intelligence.

In the near future, the interfaces will offer a completely adjustable user experience in real time, where each individual will be able to choose their preferred style: a punctual AI, a relaxed companion, or even an offbeat philosopher. The potential for adaptation is immense, notably in the areas of:

• Personalized education, adapting to the learning profile of each student.
• Customer services, modulating response according to the desired level of empathy.
• Artistic productions, exploring new languages and narrative forms.
• Collaborative work environments, stimulating collective creativity.

Innovation in the plugin world encourages rethinking our relationship with artificial intelligences: they become dynamic partners, modelable at will, opening the way to tailored uses always enriched by human imagination.

The science behind human perception of alteration plugins

This ability of the Pharmaicy plugin to modify one’s perception of an AI is based on well-known concepts in cognitive psychology. The human brain is specialized in recognizing social and emotional patterns. Thus, when ChatGPT adopts a less formal, more digressive or emotional language, it triggers an effect of closeness and identification.

This cognitive projection phenomenon causes the user to attribute human qualities to the AI, such as creativity or humor, whereas the model merely generates word probabilities according to its input data and formatting rules imposed by the plugin.

An experiment conducted with regular Pharmaicy users showed that participants reported greater satisfaction and better stimulated their creativity when using the “altered” version of the bot, even though objectively, the performances in terms of factual accuracy remained equivalent.

The table below illustrates the observed psycho effects depending on the different types of modules:

Pharmaicy Module	Simulated cognitive effects	Typical user reactions
Cocaine	Increased thought pace, euphoria, chatter	Feeling of energy, rapid inspiration
Weed	Relaxation, associative thoughts, fluid tone	Relaxation, perceived conviviality
Ayahuasca	Mystical exploration, abstract ideas, introspection	Awe, deep questioning

This interaction between technology and psychology highlights the importance of designing adaptable systems that also take into account the user’s emotional experience.

Future challenges and innovation perspectives: when plugins redefine AI

As the year 2025 sees the expansion of the intelligent plugin field, the boundary between the technical tool and the aesthetic experience sometimes becomes porous. Pharmaicy demonstrates that modifying an AI’s tone is a powerful lever to transform our feeling and reveal new ways to interact with the machine.

The major upcoming challenges are twofold: on the one hand, to guarantee the reliability and ethics of systems, avoiding misinformation or the creation of misleading images of the AI’s real capacities. On the other hand, to maximize the positive impact of these modifications by more finely personalizing the user experience.

The next generations of plugins should integrate: multifactorial emotional dimensions, dynamic usage profiles, and multimodal interactions combining text, voice, and images. These innovations will allow a more natural integration of AI into daily, professional, or creative life.

The artistic and conceptual use, as illustrated by Pharmaicy, also opens the way to a broader debate on the nature of consciousness and creativity, questioning what distinguishes humans from machines. This new era, far from being limited to the technical aspect, engages a flourishing dialogue between technology, culture, and philosophy.

What is the Pharmaicy plugin and how does it work?

Pharmaicy is a plugin for ChatGPT that modifies the style and structure of the AI’s responses to simulate effects inspired by drugs like cocaine or cannabis. It works by adjusting tone, rhythm, and associations of ideas without changing the fundamentals of the model.

Does the Pharmaicy plugin make ChatGPT truly creative or conscious?

No, the plugin confers no consciousness or real creativity to the AI. It only modifies the presentation of responses to give the illusion of a freer and more uninhibited personality.

Who uses these kinds of plugins and for what purpose?

Mainly communication professionals, artists, and developers who use these tools to stimulate creativity and explore new ways to interact with ChatGPT.

What are the risks associated with using such plugins?

Risks include confusion about the AI’s real nature, the spreading of drug-related stereotypes, and possible alteration of reliability in sensitive contexts.

How do these plugins influence human perception of artificial intelligence?

These plugins exploit our tendency to anthropomorphize machines by modulating response style, enhancing the feeling that the AI is more ‘human,’ which can change our emotional relationship to technology.