People have placed a lot of hope and expectation on OpenAI, while many also feel fear and jealousy towards this Microsoft-backed company. Everyone wants to have their own 'ChatGPT moment' in the field of artificial intelligence, but no one has been able to come close to the achievements of OpenAI. Google attempted to compete with Bard, but even with active internet connections, it failed to achieve the same level of success.
Google has reintroduced DeepMind into this competition to raise the stakes. OpenAI should be nervous at the moment as their employees are leaving to join Google DeepMind. Furthermore, Demis Hassabis, the founder of DeepMind, recently claimed that the company's next LLM project will surpass ChatGPT. This significant statement refers to the Gemini project. Similar to GPT-4, this LLM is expected to have multimodal capabilities and combines the abilities of AlphaGo, the company's reinforcement learning-based system, which has planning, reasoning, and problem-solving capabilities.
So, what advantages does Gemini have over GPT?
From everything happening on Reddit, Twitter, and around Elon Musk, it is possible that the Tesla CEO might actually acquire Reddit. Interestingly, Musk plans to create a competitor to the "awakened ChatGPT" and may use data from Twitter, possibly including data from Reddit as well. This seems reasonable as Reddit's CEO, Steve Huffman, has been praising Musk's management and handling on their platform during difficult times.
That being said, Google and DeepMind have something that others don't - YouTube. While Twitter is filled with textual data, YouTube is a goldmine of visual, audio, and textual data in almost every language on Earth. Additionally, YouTube videos are not just memes or slideshows like on Twitter or Instagram, but also include many tutorials, lectures, and podcast videos.
If Google DeepMind claims that Gemini will be multimodal, then it may actually have the best understanding and generation models currently available. For DeepMind, obtaining data from YouTube is actually ideal - multimodal, multilingual, and from various regions. Furthermore, as a company based in London, DeepMind is subject to strict EU regulations on AI governance. Now, being part of Google, headquartered in the US, they can actually benefit from the legal framework of their parent company. Interestingly, in terms of legal issues, Google's privacy policy explicitly states that even if users own the copyright, the company has the right to use the videos for its own purposes.
On the other hand, while OpenAI claims that Google's Bard is trained on ChatGPT data, there are allegations that OpenAI also uses its own Whisper model to train YouTube data by converting audio into text. OpenAI's situation has changed.
OpenAI is still not afraid enough.
Recently, when Altman and Ilya Sutskever visited Israel during their ongoing world tour, they encountered a very direct and straightforward question from the audience. "What is the secret of ChatGPT?" and whether developers trying to replicate its functionality using open-source models like LLaMa, Falcon, or Vicuna can catch up with OpenAI's product, or if they should give up.
During Sam Altman and Ilya Sutskever's recent trip to Israel, there was no beating around the bush.
The first question: the intense debate between open-source and closed-source.
Next: "Can you tell us more about the underlying model before you perform a lobectomy?"
To this, Sutskever said that open-source is still far behind. Although it is inevitable that someone will replicate this technology in the coming years, OpenAI is already far ahead in the building process.
OpenAI is proud of what they have built. When Sam Altman was asked if a team could build what OpenAI has built, he said it is "impossible" to consider it. The only advantage that OpenAI has is the support and first-mover advantage from Microsoft, which Google has been trying to overcome in the past two years.
Hassabis' stance on the risks of artificial intelligence is not clear. He previously voted in favor of a pause on giant AI experiments but is now building something to compete with OpenAI. Therefore, while OpenAI is expanding its office in London to bring some tension to DeepMind, DeepMind's association with Google will cause trouble for Microsoft and OpenAI.
The AI revolution cannot reach its peak because leaders are still climbing while trying to pull each other down.
Source: https://analyticsindiamag.com/google-deepmind-will-eclipse-openai/
Alphabet Q1 paid subscribers reached 350M, fueled by YouTube & Google One. YouTube ad revenue slightly missed at $9.88B amid Premium growth. Total Q1 revenue hit $109.9B, beating forecasts.
Anthropic releases Claude connectors for Adobe, Blender, and Ableton, enabling AI workflow automation and app control. It also sponsors the Blender Development Foundation.
Apple introduces an App Store model splitting 12-month plans into monthly installments. On iOS 26.4+, it lowers upfront costs without reducing total price, adding billing tracking.
Google launches Aadhaar digital IDs in Wallet for India, enabling secure QR/NFC verification. A phased partner rollout targets key sectors, while passport-based credentials expand globally.
YouTube tests "Ask YouTube," an AI tool blending text and video for complex queries. Initially for US Premium users, it supports interactive follow-ups, expanding Google's AI search.
Google grants US DoD classified AI access, joining OpenAI and xAI in defense deals amid Anthropic's dispute. Contractual safeguards against surveillance and lethal weapons remain unenforceable.
Mistral AI launches Workflows, a Temporal-powered orchestration platform for secure, Python-driven enterprise AI automation featuring human-in-the-loop approvals.