MIT Media Lab graduates Karthik Dinakar and Birago Jones embarked on a class project in 2010 to create a tool for content moderation teams on social media platforms. Their project, which eventually became Pienso, aimed to identify concerning posts and combat cyberbullying. However, they encountered a significant hurdle when their model failed to recognize teenage slang and indirect language used by users. This experience prompted a crucial realization – data experts, not just machine learning engineers, should be involved in building these models.
This insight led Dinakar and Jones to develop point-and-click tools that allow nonexperts to construct machine learning models. Pienso now enables users to build large language models without any coding required. The applications for Pienso have evolved beyond cyberbullying detection and now encompass various domains, including detecting misinformation, human trafficking, and weapons sales.
The founders recognized the importance of empowering domain experts rather than simply democratizing AI. They collaborated with students from nearby schools in Cambridge, Massachusetts, to train their models, which turned out to be more nuanced and accurate than anything they could have developed alone.
Pienso’s impact extended beyond social media platforms. The founders leveraged their tool during the early stages of the Covid-19 pandemic, assisting experts in virology and infectious disease to analyze research articles on coronaviruses. The insights gleaned from this analysis helped the government strengthen critical supply chains for essential drugs.
Pienso offers an alternative to businesses concerned about data donation and privacy. With its user-friendly interface, the platform allows users to import, refine, analyze, and structure data in preparation for deep learning, all without writing a single line of code. Pienso’s recent partnership with GraphCore further enhances its capabilities, providing a faster and more efficient computing platform for machine learning.
Dinakar and Jones envision a future where effective AI models are developed by the individuals most familiar with the problems they seek to solve. They emphasize that no single model can meet all needs, and therefore, a collaborative approach is necessary, leveraging a variety of models tailored to specific use cases.
Pienso’s journey exemplifies the power of technology when harnessed by those who understand the data best. By empowering data experts, Pienso strives to create an AI future that is efficient, insightful, and, most importantly, beneficial to humanity.
An FAQ section based on the main topics and information presented in the article:
Q: What was Pienso originally created for?
A: Pienso was originally created as a tool for content moderation teams on social media platforms, with a focus on identifying concerning posts and combatting cyberbullying.
Q: What problem did the founders encounter during the development of Pienso?
A: The founders encountered an issue when their model failed to recognize teenage slang and indirect language used by users, highlighting the need to involve data experts in building these models, not just machine learning engineers.
Q: What can users do with Pienso now?
A: Users can now build large language models without any coding required using Pienso. Its applications have expanded to include detecting misinformation, human trafficking, weapons sales, and more.
Q: How did the founders collaborate with domain experts?
A: The founders collaborated with students from nearby schools in Cambridge, Massachusetts, to train their models, resulting in more nuanced and accurate models than what they could have developed alone.
Q: How was Pienso utilized during the Covid-19 pandemic?
A: Pienso was used to assist experts in virology and infectious disease during the early stages of the Covid-19 pandemic. It helped analyze research articles on coronaviruses, providing insights that helped strengthen critical supply chains for essential drugs.
Q: How does Pienso address concerns about data donation and privacy?
A: Pienso offers an alternative to businesses concerned about data donation and privacy. Its user-friendly interface allows users to import, refine, analyze, and structure data for deep learning without writing any code.
Q: What recent partnership has enhanced Pienso’s capabilities?
A: Pienso’s recent partnership with GraphCore has provided a faster and more efficient computing platform for machine learning.
Q: What is the vision of the founders for the future of AI models?
A: The founders envision a future where effective AI models are developed by individuals who are most familiar with the problems they aim to solve. They emphasize the need for a collaborative approach, leveraging a variety of models tailored to specific use cases.
Definitions for key terms or jargon used within the article:
1. Cyberbullying: The use of electronic communication to bully, harass, or intimidate others, typically occurring on social media platforms or other online platforms.
2. Machine learning: The field of study that gives computers the ability to learn and improve from experience without being explicitly programmed. It uses algorithms and statistical models to enable computers to perform specific tasks without explicit instructions.
3. Virology: The branch of science that deals with the study of viruses and viral diseases, including their classification, structure, and behavior.
4. Code: In this context, code refers to the programming instructions written in a specific programming language to create software or perform specific tasks.
Suggested related links:
1. Pienso website
2. GraphCore website
The source of the article is from the blog motopaddock.nl