In a blog post, Sundar Pichai, the CEO of Google, announced the launch of Bard. The chatbot, which is already being tested by a group of trusted testers, aims to provide access to a vast amount of global knowledge through a conversational interface. Utilizing a smaller version of LaMDA, a highly advanced AI model, Bard draws upon similar technology to that of ChatGPT.
Google intends to make the chatbot available to a wider audience and gather feedback in order to improve its quality and accuracy.
OpenAI’s new language model, GPT-4, is now available and can analyze images and produce human-like speech, pushing the limits of AI technology and ethics.
Both Google and OpenAI are developing their chatbots using text generation software, which is known for its fluency but can also result in misinformation and the replication of inappropriate language sourced from the internet.
Addressing the limitations.
To address these limitations, and the difficulty in updating the software with new information, building powerful and profitable products like reinventing web search through chatbots remains a challenge.
For instance, if someone asks “Is it easier to learn the piano or the guitar?”, the response would be “Some argue that the piano is easier to learn as finger and hand movements are more intuitive, while others claim that learning chords on the guitar is easier.” Pichai also stated that Google plans to make the underlying technology accessible to developers through an API, similar to what OpenAI is doing with ChatGPT, although no specific timeline was provided.
The widespread enthusiasm surrounding ChatGPT has led to speculation that Google is facing a significant challenge to its web search dominance for the first time in years.
Microsoft, who recently invested approximately $10 billion in OpenAI, is set to host a media event tomorrow related to its collaboration with ChatGPT’s creator, which is rumored to announce new features for its second-ranked search engine, Bing.
OpenAI’s CEO, Sam Altman, tweeted a photo of himself with Microsoft CEO Satya Nadella shortly after Google’s announcement.
ChatGPT, which was launched by OpenAI in November, has quickly become an internet sensation. Its ability to provide coherent and clear answers to complex questions has many users envisioning a revolution in education, business, and daily life. However, some AI experts warn of the limitations of the tool, as it does not truly understand the information it presents and is inherently prone to fabrication.
This situation may cause frustration for some of Google’s AI specialists, as the technology behind ChatGPT was developed by the company’s own researchers, a fact acknowledged by Pichai in Google’s recent blog post.
“Six years ago, we re-focused the company on AI and have since made continued investments in AI technology,” wrote Pichai. He highlighted the contributions of Google’s AI research division as well as the work at DeepMind, the UK-based AI startup that was acquired by Google in 2014.
ChatGPT is based on GPT, a transformer AI model that was first created at Google and predicts what comes next in a string of text.
OpenAI has gained recognition for showing how feeding vast amounts of data into transformer models and increasing computing power can produce systems capable of generating language or images. ChatGPT improves upon GPT by allowing humans to provide feedback to a fine-tuning AI model that adjusts the output.
Google has taken a deliberate approach in incorporating the technology behind LaMDA into its products, due to the potential for AI models trained on web-scraped text to produce incorrect information, exhibit racial and gender biases, and spread hateful language.
This was emphasized by Google researchers in a 2020 research paper, which resulted in the firing of two ethical AI researchers, Timnit Gebru and Margaret Mitchell.
However, some Google researchers who worked on LaMDA became disheartened by the company’s slow progress and decided to leave and start their own ventures utilizing the same technology. The emergence of ChatGPT has seemingly prompted Google to expedite its plans for integrating text generation capabilities into its products.