We used X application programming interface, Twitter application programming interface for academic research, and database to identify posts related to BCI using the search term “brain-computer interface.” The acronym “BCI” was not used to avoid including irrelevant posts. The search was conducted from X inception (March 2006) to May 2022, prior to Elon Musk’s acquisition of X. The data underwent preprocessing, which involved omitting any mentions, URLs, and hashtags, removing any line breaks, deleting any HTML characters, replacing them with their respective Unicode equivalent, eliminating any special characters or punctuation points except exclamation points (the only punctuation mark relevant for sentiment analysis), and excluding posts from users with fewer than 10 followers to minimize “bot” influence and duplicate entries. We excluded posts before January 2010 due to limited data availability and after December 2021 to maintain the temporal consistency of the dataset, as our data cover only a few months of 2022 (Figure S1 in Multimedia Appendix 1). Detailed individual post data included the text, date and time of posts, the number of reposts, replies, likes, and quote count. Additional data included whether the post included links, media, tagging, or any hashtags. User information included username, the number of followers, the total number of author posts, user biography, and location.

To explore users’ demographics, we used the sentiment.ai [24] library match function after preprocessing user biographies and excluding users with empty biographies. Sentiment.ai is a text-based deep machine learning tool that allows for category matching of the most similar phrase and its category, providing a cosine similarity score. The categories (ie, the biographic groups) and phrases (ie, attributes) are shown in Table S1 in Multimedia Appendix 1. We applied categories such as “broadcast,” “scientific,” “entrepreneurship,” and “clinical” based on predefined attributes. We measured the cosine similarity score between profile biographies and the most similar attribute, categorizing them as “others” if the similarity score is less than 0.05 [23] (Figure S2 in Multimedia Appendix 1).

Changes in sentiments were tracked over time. The Mann-Kendall trend test was applied to identify significant trends in monthly incidence. Temporally weighted analyses were used to identify any changes in users’ discussions and comments regarding BCI technologies over time.

To better understand the context of topics within the complex discussion regarding BCI in X, we used the BERTopic tool [28] to conduct a topic modeling analysis. BERTopic is a topic modeling technique that uses deep learning and NLP to better understand the context and semantic relationships within text data generating clusters of similar texts, interpreted as topics. It works by embedding text into numerical representations, which are converted into a high-dimensional vector using a pretrained transformer model. Text or sentences related to each other will be close to each other in this vector, while unrelated ones will be away from each other. After dimensionality reduction of the vector, the lower-dimensional embeddings are clustered to group similar documents together. Each cluster represents a potential topic. Finally, each topic is represented by its most representative documents or by extracting keywords that best describe the cluster. Results from this topic modeling analysis include the frequency of certain words within a topic and the probability that given words represent certain topics. We preprocessed the data by removing duplicate posts, then used the “all-MiniLM-L6-v2” embedding model and the SentenceTransformer model to analyze topics in the posts discussing BCI. The top 8 topics discussed in our dataset are visualized in topic word score bar charts, including specific topics for posts expressing positive or negative sentiments alone (Figure S4 in Multimedia Appendix 1)

All analyses were conducted using Python (version 3.8; Python Software Foundation) and R (version 4.4.0; R Foundation for Statistical Computing). The P value threshold for statistical significance in this study is set at .05.

All extracted data used and presented in this study were archival, cross-sectional, and observational, obtained from publicly accessible sources without any interaction with social media users and with their usernames omitted. As such, institutional review board approval was not required.

Our study analyzed a total of 65,340 posts, created by 38,962 distinct users (Table 1). These users had a median follower count of 662 (IQR 173-2332). The median number of total posts per user was 8976.5 (IQR 2294-33,677). Most of the posts came from a diverse user base, as only 5.21% (3405/65,340) originated from the top 50 most active contributors. The content of these posts varied, as 60,079 (91.94%) included links, while 5838 (8.93%) featured media. A total of 16,617 (25.43%) posts contained tags, and 18,623 (28.5%) posts contained hashtags. Engagement metrics showed that 17,141 (26.23%) posts received at least 1 like, 5104 (7.81%) posts had at least 1 reply, 12,210 (18.68%) posts were reposted at least once, and 2688 (4.11%) posts contained quote posts. Post characteristics from 1000 randomly selected posts and a text duplicate–free sample are showcased in Table 1.

Biographic analysis (unique users=34,565) showed that “broadcasting” (10,171/34,565, 29.42%) was the highest group, followed by “entrepreneurship” (9359/34,565, 27.02%), “scientific” (9200/34,565, 26.61%), “clinical” (4066/34,565, 11.76%), and “other” (1769/34,565, 5.11%). The contribution to the BCI discussion (total posts=58,030) was the highest from users in the “broadcasting” group with 17,803 (30.67%) posts, followed by “scientific” with 16,005 (27.58%) posts, “entrepreneurship” with 14,008 (24.13%) posts, “clinical” with 7380 (12.71%) posts, and “other” with 2834 (4.88%) posts (Table 2). The “scientific” group had the highest engagement metrics among all the biography groups, while the “entrepreneurship” group had the lowest engagement metrics in our dataset (Table 3).

In 2017, there was a substantial increase in the number of posts, which accounted for approximately 24.52% of the entire dataset (Figure 1). March had the highest number of posts (n=3686), followed by April (n=3094) (Figure 2). This rise in March’s posts was primarily due to Elon Musk’s announcement of “Neuralink,” his BCI company [29]. The words “Neuralink,” “Musk,” and “Elon” were collectively mentioned 5831 times in March. The number of posts in April can be attributed to the announcement of Facebook’s BCI projects, with the term “Facebook” receiving 1944 mentions [30]. A bar chart containing the 25 most frequently mentioned terms, excluding BCI-related terms is included in Figure S5 in Multimedia Appendix 1.

Most of the posts tended to express neutral sentiments (38,804/65,340, 59.38%). Approximately one-third of the posts conveyed positive sentiments (21,404/65,340, 32.75%), while a smaller portion expressed negative sentiments (5132/65,340, 7.85%). Deletion of duplicate text resulted in a notable polarization of sentiments, potentially due to the exclusion of posts containing only titles of news articles, which tend to be more neutral in tone (Table 4). A considerable increase in positive posts was observed in February 2017, coinciding with the publication of Stanford-led BCI studies, and in July 2017, following the Defense Advanced Research Projects Agency (DARPA) award for BCI [31-33]. The number of posts expressing negative sentiment remained low, with an average of 428 (8.19%) posts per year throughout the study period (Figure 3; Figure S6 in Multimedia Appendix 1). The average polarity score showed an overall positive trend throughout the study (Mann-Kendall Statistic=0.266; τ=0.266; P<.001; Figure S7 in Multimedia Appendix 1). The mean sentiment score increased substantially upon the announcement of the collaboration between Stanford University and BrainGate in November 2011, in April 2016 following the publication of the study on the use of BCI in restoring functional movement in apatient with quadriplegia, and again in July 2017 with the DARPA-led award [33-36]. The spike in negative sentiment posts in March 2017 was mostly due to discussion related to Musk’s involvement in BCI, with an excited yet conservative tone. The word “help” (n=1489 mentions) was common in posts expressing positive sentiment, whereas in posts with negative sentiment, the term “injury” (n=325 mentions) was the most mentioned word. The frequent mentions of Elon Musk and artificial intelligence (AI) in the positive and negative sentiments are suggestive of mixed and polarized opinions about him and his BCI company “Neuralink” (Figure S8 in Multimedia Appendix 1). Most of the posts were objective (50,847/65,340, 77.81%), while fewer were subjective (14,393/65,340, 22.02%; Table 4).

The most prevalent emotion scores observed were “anticipation,” “trust,” and “fear,” accounting for 20.52% (10,802/52,618), 17.56% (9244/52,618), and 13.95% (7344/52,618) of the expressions, respectively. However, “surprise,” “joy,” and “disgust” were less frequently expressed, constituting 8.77% (4619/52,618), 12.69% (6681/52,618), and 2.79% (1470/52,618), respectively (Table 5). The emotion score of the text duplicate–free dataset and random 1000 posts are in Table 5. Focusing on the 3 most prominent emotions, “anticipation” peaked at 47.49% in 2010, coinciding with the sharing of news regarding the potential gamification of BCI technologies [37]. In July 2017, there was a substantial increase in posts expressing anticipation as their primary emotion reaching as high as 48.07% of posts shared, discussions were mostly related to DARPA’s award for BCI, and the use of BCI in the production of music using the “encephalophone” [33,38,39]. “Trust” reached 23.43% of total posts in 2021 and was the lowest in 2016 with 3.57% of total posts. “Fear” displayed a notable increase in posts in 2016, with 25.90%, and was the lowest in 2013 with 3.34% (Figure 4; Figure S9 in Multimedia Appendix 1).

A temporally weighted analysis was conducted to explore the change in public discussion before, during, and after the spike in 2017 (Table 6). The increase in posts expressing a neutral sentiment (11,394/16,023, 71.11%), as well as the rise in posts expressing anticipation (1417/2512, 56.41%) and trust (n=1117/2512, 44.47%) in 2017, demonstrates a public that is excited yet cautious toward this emerging technology. However, after 2017, the public discussion polarized, as positive and negative views increased substantially, from 24.41% (3912/16,023) to 40.86% (10,391/25,430) and from 4.47% (717/16,023) to 9.41% (2393/25,430), respectively. Additionally, there was a notable increase in subjective posts, from 18.95% (3036/16,023) to 26.76% (6804/25,430), after 2017. Fear was substantially expressed before 2017 with 40.97% (1936/4725), then dropping to below 30% during and after 2017, possibly indicating the public excitement, and trust following the big announcements in the BCI industry in 2017. However, the increase from 23.01% (578/2512) to 28.3% (2237/7904) in fear might indicate unaddressed ethical questions.

We conducted a topic modeling analysis to understand the semantic and dynamic discussion related to BCI technology. The main topics discussed in our dataset included ideas related to BCI changing the future, Musk’s announcement of “Neuralink,” Facebook’s involvement in BCI technology, and the gamification of BCI technologies (Figure S10 in Multimedia Appendix 1). Topics discussed in posts expressing positive sentiment included the advancement of BCI in research and health care and BCI advancements with virtual reality (VR) and augmented reality (AR; Figure S11 in Multimedia Appendix 1). Discussions in posts conveying negative sentiment were related to conspiracy theories of Musk or “Neuralink,” further demonstrating that the involvement of Musk in BCI was although highly influential but also particularly controversial. Moreover, our analysis detected discussions related to the potential use of BCI for malicious reasons including mind control (Figure S12 in Multimedia Appendix 1). Examples of posts related to different topics are included in Table S2 in Multimedia Appendix 1.

Our study highlights the complex and evolving public perception of BCI technology, as expressed on X discussions over the past decade. Sentiment analysis reveals that while a majority of the public holds neutral views, there is a notable undercurrent of anticipation, trust, and fear. The presence of these conflicting emotions suggests that while there is optimism about the potential of BCIs, concerns around ethical issues remain unresolved. Understanding these emotions is essential for the development and acceptance of BCI technology. Key topics that emerged from the discussions include Neuralink and Elon Musk—subjects of both positive and negative conversations—the practical applications of BCIs, and the potential for gamification of BCI technology. In addition to being key market drivers, addressing these factors will influence innovation and drive investments in the BCI space.

BCI is one of the most important technologies in neuroscience, sparking diverse opinions regarding its ethical implications [40-42]. Safety, justice, privacy, and security are the main concerns discussed in ethical literature [43]. Some authors had concerns over the safety of implanting such invasive technologies, which could potentially cause serious health complications [44], while others have criticized the biased narrative in the BCI literature, which tends to overlook the perspectives of individuals with disabilities [45]. There was unease about the handling and storage of sensitive data obtained through BCI, with worries about its potential exploitation for malicious purposes [46]. In our study, we detected discussions reflecting many of these concerns. We specifically highlighted concerns regarding Elon Musk and his company Neuralink, particularly related to safety and animal rights issues [47], as well as the potential for BCI to be used in terrorism, or for brain hacking, which could lead to the leaking of sensitive personal information or even revealing emotions and thoughts. Open and inclusive discussions are essential to guiding the ethical development and use of BCI technology. Moreover, users from the scientific community will play a particularly critical role in this process. Our study found that posts from scientists generated the highest engagement when discussing BCI, underscoring their influence in shaping public understanding. By leveraging their trusted position, scientists can educate and foster informed discourse on these ethical issues to steer the responsible development and application of BCI technology.

The significant rise in digital discussions about BCI in X, particularly in March 2017, aligns with Elon Musk’s announcement of his BCI company, Neuralink. Elon Musk’s substantial social media influence is evident, with almost 200 million followers on X. His posts not only are influential but often polarize public opinion [48]. For instance, his comments on cryptocurrencies have demonstrably impacted their market values [49]. Additionally, a sentiment analysis focused on electric vehicles revealed that Musk was a central figure in these discussions [50]. These observations reinforce the significant role of influential figures like Elon Musk in steering the public dialogue and development of cutting-edge technologies.

Most posts conveyed a neutral sentiment, reflecting a combination of anticipation and excitement, yet mixed with fear and anxiety, resulting in a generally doubtful public perception. This perception could potentially be the result of unanswered ethical challenges. Furthermore, Musk’s involvement in BCI technologies has significantly influenced public perceptions, often polarizing opinions. The 2023 sentiment analysis conducted on Reddit indicates that neutral and negative views toward Elon Musk outnumber positive sentiments in discussions about him [48]. Additionally, in our study, a common theme in both the negative and positive discussions about BCI frequently centered around Musk or Neuralink. This mixed perception of Elon Musk could be another contributing factor to the sentiment results of our study. Ensuring transparent communication from influential figures like Musk might be key to positively shifting public sentiment toward BCI technologies.

Since acquiring Twitter, now rebranded as X, Elon Musk has become actively involved in geopolitical discourse, leading to conflicting opinions about his involvement in politics. This increased political engagement has further polarized public perception of him [51,52], which might potentially extend to BCI technologies. To mitigate this migration of polarization, it is essential to diversify the voices in BCI communication by encouraging scientists and industry experts to take a more prominent role, thereby reducing the focus on any single individual.

Nonetheless, Musk’s involvement in BCI technology has had a significant and positive impact, contributing to greater public awareness and advancements in the field. The surge in BCI-related discussions in 2017, following Musk’s announcement of Neuralink, not only increased awareness of the technology and its potential applications but also led to a rise in posts expressing positive sentiments, mixed with anticipation and trust. This suggests a general optimism regarding the potential of BCI technology. Furthermore, this trust and enthusiasm is reflected in the recent growth of BCI-related publications as well as the rapid expansion of the BCI market [53,54]. In summary, even though Musk is considered a controversial figure by many, his influence in the field of BCI has been largely positive. However, it is important to continue making conscious efforts to ensure that this positive impact is sustained and that the technology develops responsibly and ethically.

We found that anticipation and trust were the most expressed emotions, likely stemming from the involvement of renowned universities in BCI research, including the collaborative BrainGate2 project [34,35]. The widespread media coverage and remarkable BCI outcomes in individuals with severe disabilities have also contributed to this trend [31,32,36]. These technologies enable patients with disability to overcome their disabilities, thereby enhancing their quality of life. The public enthusiasm observed toward these milestones underscores a substantial societal demand for such innovations. This heightened positive sentiment indicates that the integration of BCIs into neuroprosthetics, AI-driven prosthetics, and exoskeleton devices not only represents a significant technological advancement but also fulfills a critical public need for effective solutions to address disabilities. Moreover, BCI technology used for nonrehabilitation activities, such as gaming and music production, led to an increase in anticipation due to its potential appeal and application to everyday consumers.

Fear was the third most prevalent emotion identified. This may be related to ongoing unanswered ethical concerns in BCI development, varying public opinions about public figures, and inherent fear toward novel technologies. Our capture of fear might be an early indicator of potential technophobia, as BCI becomes more commercially available. Difficult to understand and complex technologies are associated with more anxiety and fear of using them [55]. In addition, the concept of “mind control,” frequently portrayed in popular culture and science fiction, could potentially heighten fear perceptions. Upcoming depictions and future works that continue to explore this theme may trigger more apprehensive emotions. Such emotion, if not addressed, could foster a conservative mindset, potentially slowing the adoption and hindering the application of innovative technologies such as BCI.

One particularly exciting avenue for BCI technology is its integration into gaming. The potential of gamification in BCI technologies is vast, offering applications not only for medical rehabilitation [56,57] but also for entertainment experiences [58-60]. The concept of controlling and fully immersing oneself in a game to create experiences beyond physical limitations is highly appealing to the public, as evidenced by a notable increase in anticipation. Gabe Newell, cofounder and president of Valve, showcased Valve’s interest in BCI for gaming, sparking significant enthusiasm [61]. Integrating BCI technologies into gaming could play an important role in further advancing their development, much like the early adoption of VR and AR through gaming platforms such as the HTC Vive (by HTC and Valve Corporation), Meta Quest (by Meta Platforms), and PlayStation VR (by Sony Interactive Entertainment). These technologies, which began in gaming, have since expanded into broader applications in healthcare and education.[62]. Fueled by public enthusiasm, several companies have begun claiming that their products leverage BCI to enhance gaming experiences. However, it remains debatable whether we have truly reached that point [11]. Numerous attempts have been made to use BCI for gaming, using both invasive and minimally invasive technologies. For example, Neuralink has successfully implemented a BCI in a human with quadriplegia, allowing the individual to play chess and other games by moving a cursor [63]. “I basically have an aimbot in my head,” said Noland Arbaugh, the first patient to receive a brain-computer chip from Neuralink. An “aimbot” is a type of cheating software used in video games that automatically aims at opponents at a superhuman speed [64,65]. Similarly, a popular Twitch streamer successfully beat the notoriously difficult game Elden Ring using a wearable electroencephalograpy-based BCI device, with thousands of live viewers [66]. With a global video game market size estimated at US $217.06 billion in 2022, these advancements are likely to positively influence public perception of BCI and may attract the attention of major players in the gaming industry, potentially driving further research and development of BCI technologies [67]. Gamification thus presents a significant opportunity for the future of BCIs, acting as a gateway for mainstream consumer adoption, similar to how VR or AR technologies gained traction.

The economic outlook for BCIs is promising. In 2023, the global market for BCIs was valued at US $2.0 billion and is projected to grow at a compound annual growth rate of 17.8% from 2024 to 2030. This growth is driven by increasing demand for neuroprosthetic devices and advancements in technology that enhance mobility and communication for patients who are paralyzed as well as expanding applications in gaming and military communication [53]. However, public opinion and acceptance are essential to sustaining this market growth. Greater acceptance of the technology will drive further adoption and development, especially since young investors and future philanthropists are likely to support companies that align with their personal values and contribute to humanitarian goals [68]. Our findings align with this broader market analysis. Most individuals in our study exhibit a neutral sentiment toward BCI technology, reflecting a phase of critical appraisal of this emerging innovation. Additionally, the prominence of emotions such as trust, anticipation, and fear underscores the imperative to educate, demystify, and familiarize the public with BCI technology. Doing so will be key to enhancing acceptance among the public. Additionally, the increase in the number of subjective posts in our analysis reflects increased awareness of BCI as a topic for discussion, and this growing awareness suggests that the public will begin to form their own ideas, opinions, and emotions toward BCI. This awareness might also indicate that people are shifting from passive observers to active participants in BCI discussions, contributing to the evolving conversation and potentially influencing its trajectory. Therefore, a concerted effort to inform and engage the public is crucial for fostering positive sentiment and ensuring the responsible development and integration of BCI technology. This, in turn, will support the continued expansion of the global BCI market.

The study has several limitations that may have impacted the accuracy of our results. First, the search excluded the term “BCI” to avoid including irrelevant posts, which may have inadvertently omitted relevant content. Additionally, the analysis was restricted to English-language posts and focused solely on the platform X, potentially overlooking important discussions on other platforms or in different languages. It was not possible to infer some of the important demographics of users missing out on important contextual information. We used existing tools for NLP analyses that were not tailored and validated specifically for the study’s topic, which may potentially lead to inaccurate results. Determining the geographical locations of users was not feasible, which would have provided valuable regional insights. Regarding the biography analysis, some overlap between categories is expected, as a single user may fall into multiple classifications. This method of short bioclassification is effective for only a subset of users, and further validation is required in future studies. Finally, the substantial number of posts containing links may indicate that the sample is not representative of general discussions but primarily consists of shared news articles. Future research should address these limitations to enhance the depth and accuracy of the understanding of the public perception toward BCI technologies.

The findings from this NLP-assisted study offer a decade-long overview of public perception of BCI technology. Overall sentiment was mostly neutral, but the emotions most commonly linked to BCI—anticipation, trust, and fear—reflect a complex emotional response, suggesting a cautiously optimistic yet apprehensive attitude toward the advancement of BCI technology. Notably, the presence of fear underscores the importance of addressing ethical concerns and ensuring transparent communication within the BCI field. Resolving these issues is critical for reassuring the public and mitigating apprehensive attitudes. Additionally, the involvement of influential figures and leading institutions, along with reports of positive clinical outcomes, such as advancements in neuroprosthetics and rehabilitation, may foster more favorable public perceptions of BCI technology. The gamification of BCI, particularly its integration into gaming and entertainment, also offers a pathway to increased public engagement and adoption. However, the persistence of fear signals potential resistance that may impede progress if left unaddressed. Prioritizing ethical transparency, expanding public education, and incorporating a more diverse range of voices in the discussion could help drive broader acceptance and responsible use of BCI. Although this study uses advanced AI tools to offer valuable insights into public sentiment, certain limitations, such as potential biases and incomplete demographic data, should be acknowledged. Nonetheless, the findings serve as a valuable reference point for guiding future technological developments, informing policy making, and crafting effective communication strategies within the BCI sector.