Despite Google’s significant advantages in terms of cloud computing resources and human talent, it is frequently perceived as lagging behind in the generative AI competition. This view was reinforced when Microsoft CEO Satya Nadella stated that Microsoft had likely already missed its chance to excel in artificial intelligence. In response, Alphabet’s CEO Sundar Pichai made sharp remarks about Microsoft’s faltering collaboration with OpenAI.
I’m always ready for a direct comparison between their models and ours; it doesn’t matter when, just whenever. It seems they are utilizing other models at the moment.
However, it appears that the situation has taken an ironic twist; Google, who was seemingly criticizing another entity for using similar technology, might have been doing the same. According to exclusive documents reviewed by Business Insider, Google’s contractors at Scale AI reportedly employed ChatGPT as a tool to enhance and develop Bard (now Google Gemini).
The damning report also showed that contractors compared a vast number of responses produced by ChatGPT with those generated by Bard. These comparisons were then utilized to enhance Bard’s responses, ensuring they were at the very least equivalent to those from ChatGPT.
It’s true that Satya Nadella stated that OpenAI has about two years to advance and perfect ChatGPT without much competition, which could pose a challenge for rival models such as Copilot and Gemini in trying to keep pace.
AI managers at Scale recognized that OpenAI’s chatbot was producing more efficient, well-structured replies enriched with intriguing details. The task set before contractors was to enhance Bard beyond GPT-4 performance level, and a 15% bonus incentive was offered for responses that significantly surpassed ChatGPT in quality.
It’s important to point out that according to OpenAI’s user agreement, their output cannot be utilized to educate competing models. Yet, Scale AI has refuted these allegations.
“Scale did not, and does not, use ChatGPT responses to train Gemini or any models.”
According to Scale AI, the documents acquired by the media outlet showcase typical comparative assessments, a type that’s frequently misunderstood as real-world application of ChatGPT in the process of model building and enhancement. The AI company clarified that these evaluations are consistent with standard industry practices.
Meta’s new $14.8 billion stake in Scale AI drives rift with Google
Recently, it was disclosed that Meta has agreed to buy a significant portion of Scale AI for approximately $14.3 billion, which will give them 49% ownership of the artificial intelligence company. This transaction boosted Scale AI’s market value to an impressive $29 billion. Additionally, it’s important to mention that the creators of Facebook have also recruited Scale AI’s founder, Alexandr Wang, to head up their newly established unit focused on advanced artificial intelligence.
It seems that Meta’s recent partnership with Scale AI is causing tension with Google. According to Reuters, Google may choose to end its relationship with Scale AI, their largest customer. Google was set to pay up to $200 million to Scale AI in 2025 for labeled training data, a crucial component in the creation of AI models used by Gemini.
Read More
- Forza Horizon 5 Update Available Now, Includes Several PS5-Specific Fixes
- Gold Rate Forecast
- ‘The budget card to beat right now’ — Radeon RX 9060 XT reviews are in, and it looks like a win for AMD
- Masters Toronto 2025: Everything You Need to Know
- We Loved Both of These Classic Sci-Fi Films (But They’re Pretty Much the Same Movie)
- Valorant Champions 2025: Paris Set to Host Esports’ Premier Event Across Two Iconic Venues
- Karate Kid: Legends Hits Important Global Box Office Milestone, Showing Promise Despite 59% RT Score
- Eddie Murphy Reveals the Role That Defines His Hollywood Career
- Discover the New Psion Subclasses in D&D’s Latest Unearthed Arcana!
- Street Fighter 6 Game-Key Card on Switch 2 is Considered to be a Digital Copy by Capcom
2025-06-17 00:39