① The successive releases of OpenAI's GPT-4O and Google's Astra show that technology companies all attach great importance to the development of artificial intelligence assistants; ② Judging from current use cases, they cannot be said to be essential products for everyday life, but future growth is likely to make them a “killer app”
Financial Services Association, May 15 (Editor Zhou Ziyi) This week, the headlines in the field of artificial intelligence are undoubtedly the product duel between OpenAI and Google.
OpenAI has always “loved” releasing its products before competitors' major product launches to seize the spotlight, and this week was no exception.
OpenAI had previously given high expectations to the public. On Monday (May 13), the company announced an upgraded version of GPT-4 as scheduled, called GPT-4o (“o” represents omni omni-omnidirectional). GPT-4o is designed to act as a personal assistant on a phone or tablet, with improved voice interaction capabilities, the ability to interpret and reason pictures taken by the device's camera, more powerful language translation capabilities, and faster response times.
The technical innovations behind GPT-4o are impressive. The model is multi-modal, and it can receive audio, vision, and text in real time, and generate any combination of text, audio, and image outputs. Compared to previous versions, this model eliminates the steps of converting the user's voice into text and processing it, which means the entire process is much faster.
GPT-4O also shortens the time it takes for the model to process a specific number of tokens (in the case of English text, one token is usually equal to one and a half words), which also makes the model run faster and cheaper than OpenAI's previous best model GPT-4 Turbo.
On Tuesday (5/14), Google also stepped up its moves and toughened OpenAI head-on.
At Google's I/O developer conference, Google announced a range of new artificial intelligence features and upcoming products, including extensive upgrades to the Gemini model, the future artificial intelligence assistant “Astra”, generative artificial intelligence to enable Google search, and a series of generative AI tools related to images, music, and video.
At the conference, Google announced improvements to the Gemini 1.5 Pro model, further expanding the context window of 1 million tokens to 2 million, and making it possible to have more natural sound, better understanding audio and images, stronger logical reasoning and planning capabilities, and better computer code generation capabilities.
Furthermore, Google also released Astra, an advanced visual and conversational response agent project to process multi-modal input content such as audio and video. Compared to OpenAI's GPT-4O, which can only process still images, Astra can also process video. In a demo video, it can recognize instructions such as “what makes a sound” and “where you are now” through camera video. However, there are delays or delays in its response. It is reported that future versions of Google's artificial intelligence personal assistants are being developed through “Astra.”
“Highlight Moments” for AI Assistants
As can be seen from the product launches of OpenAI and Google, technology companies all attach great importance to the development of artificial intelligence assistants, and the position of “the first killer application of artificial intelligence” has become a “must-compete place” for all companies in Silicon Valley.
Judging from this week's product launches, OpenAI and Google's artificial intelligence assistants each have advantages. GPT-4o can directly receive and generate speech, eliminating the process of converting speech into text; Astra can process moving images such as video, which is a significant advantage.
The launch of these two products clearly puts the other Silicon Valley giants Apple and Amazon at a disadvantage. They need to upgrade their voice assistants Siri and Alexa to keep up with the capabilities of these new competitors, otherwise these products will be in trouble. As far as is known, Anthropic, which Amazon invests in, has a powerful Claude AI model that can be used; previous reports also surfaced that Apple is in negotiations with OpenAI to obtain a license for its technology in the short term.
But will these new AI assistants be future “AI killer apps”? This conclusion is still inconclusive and depends entirely on what happens next.
Judging from the current use cases of artificial intelligence assistants, they cannot be called essential products ubiquitous in human daily life. Other than the translation function, almost none of them can help people complete their work.
Some analysts suggest that this may change when these assistants have more “proxy” properties. If one day they can truly understand human preferences, complete tasks according to people's preferences, and help with some things in everyday life (such as online shopping, filling out insurance forms, booking vacations, etc.), artificial intelligence assistants are likely to become a “killer app.”
Google says it is currently developing such a product, but has not given a timeline for product launch; OpenAI also continues to reveal that it is “about to release” exciting future announcements; next week, Microsoft will hold a Build developer conference.