From OpenAI's GPT-4o to Google's Gemini: All the major AI updates to know about this week
This week was all about artificial intelligence (AI) competitors show‐ ing off their might with OpenAI and Google’s DeepMind trying to trump the other.
But the impressive updates show how the technology is being rolled out quickly, living up to previously voiced fears penned in an open let‐ ter that AI could harm society and should be put on pause until it is better understood.
OpenAI’s co-founder and chief sci‐ entist who was known to question Sam Altman’s ambition to deploy AI quickly, an‐ nounced on Wednesday he was leaving the start-up.
On Monday, OpenAI held its firstever live session and announced its new model GPT-4o, which claims to be “much faster” and improves “ca‐ pabilities across text, vision, and audio”.
A version of it will be free to users, but there is a paid version that will have five times the capacity limits. The most interesting development from the company is the voice as‐ sistant that can generate content or understand commands in voice, text, or image and do live transla‐ tions. It can also respond in realtime and observe body motion.
But OpenAI did not stop there with announcements. On Thursday, the company said it had signed a deal for access to content from Reddit and would bring AI-powered fea‐ tures to the social media platform. The same day, Sony sent letters to OpenAI, Google, and Microsoft, asking if they had used its songs to develop their AI systems. The let‐ ters come as copyright cases against generative AI companies mount.
Google vs OpenAI rivalry
Google similarly signed a deal with Reddit earlier this year for its AI training data.
Competition between the two com‐ panies heated up this week as Google’s DeepMind released up‐ dates to its chatbot Gemini, a day after the OpenAI event, which ap‐ peared to be no coincidence. At Google’s I/O developer confer‐ ence, a video was shared showing that the company's Gemini chatbot was watching GPT-4o.
When asked by voice command, Google’s chatbot was able to at‐ tribute quotes to the OpenAI speakers and even named them when it was asked to. The two com‐ panies showed how voice input is overtaking text input.
Gemini can also use personal infor‐ mation, such as holiday time re‐ quests and use public information to quickly build itineraries.
But that was not all Google an‐ nounced. Its search engine is also getting a makeover. Instead of of‐ fering the Search Generative Expe‐ rience (SGE), AI will generate a summary of their search results.
Not just about OpenAI and Google
While the two AI titans appear to be leading the sector, Anthropic al‐ so had announcements this week. On Tuesday, Anthropic’s artificial intelligence (AI) assistant Claude became available in Europe. Anthropic bills itself as an AI safety research company and was founded by former OpenAI execu‐ tives and researchers. Google and Amazon are also major Anthropic investors.
“We're incredibly excited to launch in Europe because the region is well-positioned to harness the ben‐ efits of AI,” an Anthropic spokesperson told Euronews Next. “Claude has a number of specific features that uphold EU values and caters for the EU market; we de‐ signed Claude to avoid biases, dis‐ crimination, and hate speech; fo‐ cused on reliable, steerable, and accessible AI for all”.