Yann LeCun Explains How Meta Had Developed A ChatGPT-Like System Before OpenAI, But Took It Down

Yann LeCun Explains How Meta Had Developed A ChatGPT-Like System Before OpenAI, But Took It Down

  • 24.03.2025 13:18
  • officechai.com
  • Keywords: danger, success

Yann LeCun revealed that Meta developed a ChatGPT-like system called Galactica before OpenAI's release but shut it down due to public backlash. Despite this, ChatGPT gained widespread popularity, highlighting differences in public perception and readiness for AI technologies.

Meta NewsMETAsentiment_dissatisfied
open_in_newOriginal article

Estimated market influence

Meta

Meta

Negativesentiment_dissatisfied
Analyst rating: Strong buy

Galactica was developed by Meta but faced backlash and was taken down.

OpenAI

Positivesentiment_satisfied
Analyst rating: N/A

ChatGPT led to OpenAI's success in the AI race.

Context

Analysis of Meta's Galactica and Market Implications

Key Developments

  • Meta's Galactica:

    • Developed in fall 2022 by a small team at Meta.
    • Released on November 15, 2022, two weeks before ChatGPT.
    • Trained on 48 million examples of scientific articles, websites, textbooks, lecture notes, and encyclopedias.
    • Designed to assist researchers and students with tasks like summarizing papers, solving math problems, generating Wiki articles, writing scientific code, annotating molecules, and proteins.

  • Public Reception:

    • Criticized for "hallucinating" fake papers and authoritative-sounding but incorrect information.
    • Scientific community expressed concerns about reliance on unverifiable outputs.
    • Meta took the demo offline on November 17, 2022, due to backlash.

Market Dynamics

  • ChatGPT's Success:
    • Released on November 30, 2022, as a general-purpose LLM.
    • Received widespread enthusiasm and acclaim for its conversational capabilities.
    • OpenAI gained significant attention and positioned itself at the forefront of the AI race.

Competitive Landscape

  • Meta vs. OpenAI:
    • Meta's Galactica targeted the scientific community, which had high expectations for accuracy.
    • ChatGPT appealed to a broader audience, including non-experts who were less critical of its flaws.
    • Timing and perception played significant roles in the differing receptions.

Strategic Implications

  • Public Readiness:

    • The scientific community was more skeptical due to higher standards for accuracy.
    • General public showed greater tolerance for imperfections in ChatGPT, driven by novelty and excitement.

  • Regulatory Considerations:

    • Potential future regulatory scrutiny over AI reliability and misuse.
    • Companies may need to balance innovation with transparency and accountability.

Long-Term Effects

  • Impact on AI Development:

    • Highlights the importance of user perception and public relations in AI releases.
    • Emphasizes the need for clear communication about capabilities, limitations, and ethical considerations.

  • Industry Leadership:

    • OpenAI's bold move to release ChatGPT publicly established it as a leader in the AI race.
    • Meta's experience with Galactica underscores the risks of premature or misaligned product launches.

Conclusion

The contrasting fates of Galactica and ChatGPT reveal critical insights into market dynamics, user perception, and strategic decision-making in AI development. While Meta demonstrated technical prowess, OpenAI's timing, branding, and alignment with public enthusiasm proved pivotal in capturing the spotlight.