Bill Gates already said 2 years ago that ChatGPT-5 wouldn't be a revolution. „We have reached a plateau”
The failed launch of ChatGPT-5 has brought to mind Bill Gates' words from 2 years ago. Even then, the Microsoft founder suggested that AI models might be reaching their limit.
The release of ChatGPT-5 was met with negative reception from the community. OpenAI announced a revolution that would leave previous and competitive models far behind. Reality turned out to be different and it seems that it was predicted much earlier. One of the people who was skeptical about drastic changes in models was Bill Gates (via Windows Central).
Bill Gates believed that AI is reaching its limits
In October 2023, Bill Gates gave an interview to Handelsblatt. During the conversation, he stated that "we have reached a plateau" of AI development, and the successor to ChatGPT-4 will not be significantly better than its predecessor. The leap from ChatGPT-2 to ChatGPT-4 is no longer possible. According to Gates, appropriate research methods are needed for further progress, which will allow us to better understand how AI works.
Now this opinion seems to be shared by industry media, and, for example, MIT Technology Review points out that although there is an improvement compared to previous models, it is more refined than revolutionary, contrary to what Sam Altman had predicted. However, we could have seen this coming, as there were reports of tech companies facing issues as early as 2024.
The troubles concerned the construction of more advanced models of artificial intelligence, as reported by Bloomberg in November 2024. According to the information provided, the model didn't meet OpenAI's expectations during tests answering questions about coding, for which it had not been previously trained. Similar problems were also said to concern models from Google (Gemini AI) and Anthropic (Claude AI).
For now, it seems that Bill Gates' words are proving true in practice. Corporations are struggling to achieve the satisfying results that would confirm the promises of the AI revolution. Despite grand declarations about the impending AGI, the slowing progress seems to push back the vision of truly conscious artificial intelligence in time.
- GTA's success „brought the US government down on us,” says ex-Rockstar co-founder. „Some people quit, it was really tough”
- OpenAI CEO declares „code red” because of Google. Billion-dollar compensation payout is looming over the company
- „We won't use AI because we respect the actors.” James Cameron assures that we won't see artificial intelligence in Avatar
0
Author: Zbigniew Woznicki
He began his adventure with journalism and writing on the Allegro website, where he published news related to games, technology, and social media. He soon appeared on Gamepressure and Filmomaniak, writing about news related to the film industry. Despite being a huge fan of various TV series, his heart belongs to games of all kinds. He isn't afraid of any genre, and the adventure with Tibia taught him that sky and music in games are completely unnecessary. Years ago, he shared his experiences, moderating the forum of mmorpg.org.pl. Loves to complain, but of course constructively and in moderation.
Latest News
- Now you can get Bobcat blueprint in ARC Raiders easily. Here's what you have to do
- Assassin's Creed meets Avatar: The Last Airbender: The God Slayer looks like an incredible open-world action adventure RPG
- ChatGPT gladly shoots a YouTuber, overriding safety protocols with minimal manipulation
- Sony teams up with Left 4 Dead creator on new 4-player cooperative shooter
- „The rebirth of historical Total War” Total War: Medieval 3 announced