Duolingo -- the next chapter in human computation | Luis von Ahn | TEDxCMU 2011

Summary

In this engaging talk, the inventor of CAPTCHA discusses its purpose in distinguishing humans from bots on the web. She then introduces ReCAPTCHA, a system that not only verifies humanity but also harnesses the time spent solving CAPTCHAs to digitize books. The speaker shares amusing anecdotes and highlights the massive scale of participation, with over 750 million people contributing to knowledge digitization. She also teases an upcoming project, Duolingo, which aims to translate the web into major languages using a novel approach that combines language learning with community-driven translation efforts.

Takeaways

• 😀 The speaker, Anny Chung, is the inventor of CAPTCHA, a system designed to differentiate human users from bots on the internet.
• 😅 CAPTCHAs can be annoying for users but serve a crucial role in preventing automated programs from performing actions like bulk ticket purchases.
• 🔄 The speaker highlights an amusing incident where CAPTCHAs displayed coincidental words, causing unintended humor or confusion.
• 🌐 The ReCAPTCHA project is an evolution of traditional CAPTCHAs, aiming to make the user's effort productive by contributing to digitizing books.
• 📚 ReCAPTCHA works by presenting users with two words during the verification process, one known and one unknown, thus aiding in the digitization of text from old books.
• 🔢 It's estimated that around 200 million CAPTCHAs are solved daily, equating to approximately 500,000 hours of human time.
• 😲 The ReCAPTCHA system has been so successful that it has digitized the equivalent of about 2.5 million books per year through user participation.
• 😆 The randomness of word pairings in ReCAPTCHA has inadvertently led to the creation of 'Captcha Art,' a meme where people create art based on amusing word combinations.
• 🌟 Over 750 million distinct individuals have contributed to book digitization through ReCAPTCHA, showcasing the power of collective online effort.
• 🌐 The speaker's research is driven by the question of what large-scale achievements can be accomplished with the internet's ability to coordinate massive groups of people.
• 📈 The project 'Duolingo' is introduced as a platform that aims to leverage the desire of over 1.2 billion people to learn languages to translate the web for free.

Q & A

What is the purpose of CAPTCHAs?

CAPTCHAs are designed to differentiate between human users and computer programs by presenting distorted characters that are easy for humans but difficult for computers to read, ensuring that the entity filling out the form is a human and not an automated bot.

Why were CAPTCHAs created?

CAPTCHAs were created to prevent automated programs from submitting forms millions of times, such as in the case of ticket scalpers trying to buy large quantities of tickets at once.

What is the issue with the random sequence of characters in CAPTCHAs?

Sometimes the random sequence of characters in CAPTCHAs can unintentionally form words or phrases that are humorous or inappropriate, leading to user confusion or unintended messages.

How does the Recaptcha project improve upon traditional CAPTCHAs?

Recaptcha improves upon traditional CAPTCHAs by turning the time spent by users typing CAPTCHAs into productive work. Users help digitize books by typing words that computers cannot recognize, thus contributing to the digitization of human knowledge.

Why is the digitization of books important?

The digitization of books is important because it helps preserve knowledge, makes information more accessible, and can be used for research, education, and other purposes.

How does the Recaptcha system determine if a user is human?

The Recaptcha system presents two words to the user, one of which the system already knows the answer to and the other it does not. If the user types the correct word for the known one, the system assumes the user is human and gains confidence in the typing of the unknown word.

What is the significance of the number 750 million in the context of Recaptcha?

The number 750 million represents the distinct number of people who have helped digitize at least one word from a book through Recaptcha, showcasing the massive impact of collective human effort.

What is the main goal of the Duolingo project mentioned in the script?

The main goal of the Duolingo project is to engage 100 million people in translating the Web into every major language for free while they learn a new language, addressing both the lack of bilinguals and motivation for translation.

How does Duolingo plan to achieve high-quality translations without paying professional translators?

Duolingo plans to achieve high-quality translations by combining the efforts of multiple language learners. As users learn and translate, their collective work can match the quality of professional translations, leveraging the large number of users to create value.

What is the innovative business model proposed by Duolingo?

Duolingo's innovative business model allows users to learn a language for free by translating web content, creating value that can be monetized. This model does not require users to pay with money but instead with their time, making language education more accessible to a broader audience.