At its I/O developer conference on Wednesday night, Google revealed how it plans to eclipse ChatGPT without making the same mistakes. This post revolves around the major software announcements and asks what’s available in this country.
«Magic Editor» for Google Photos
Google’s latest Photos feature is a feature called Magic Editor. Generative AI makes it easy to make big changes to a shot.
According to the announcement, the AI feature for Google Photos should be available later this year – initially only on “select” new Pixel smartphones.
“Magic Compose” is intended to enhance lyrics
The new AI function “Magic Compose” should be able to be used in messages and conversations to rewrite text in different styles.
That’s how you do it: According to the announcement, to get started with Magic Compose, you’ll need to use Google’s “Messages” app and first enter the text as usual. You then choose how you want the message to sound, and the generative AI can adjust the text accordingly.
According to Google, “Magic Compose” will launch this summer as a beta version for Android users.
Google boss Pichai demonstrated, among other things, how software can formulate a letter for the user. And if you’re writing a story, the software should be able to suggest further twists in the story and auto-generated illustrations.
Google Bard gets a beef, but…
Google’s ChatGPT challenger will soon be available in 180 countries, albeit in English only. Anyone interested in testing the AI chatbot does not have to be put on a waiting list.
However, when the Watson editor logged into bard.google.com (using his own Google account), there was a bitter disappointment: “Currently unavailable” was displayed on the screen. 💩
Japanese and Korean are now also supported, but German and Swiss German are not. Another 40 languages, including German, will follow “soon”. But Bard will also not be available in the countries of the European Union (EU) for the time being.
According to the announcement, Google Bard is getting a number of new features, including easier ways to export text to Google Docs and Gmail, visual search, and a dark mode.
The new AI chatbot can not only process speech in text form, it also “understands” image content to a certain extent. This is called multimodal.
The Google Workspace apps, i.e. the office software, will also be complemented by AI features to make them more powerful: this should happen, for example, with the addition of automatic table generation (no formulas) and automated image creation in Slides and Meet.
This is what the new language model PaLM 2 offers
During the developer conference, Google announced the launch of PaLM 2, the latest Large Language Model (LLM). The new AI chatbot Bard and other AI functions are also based on this technology.
Like its competitor OpenAI is doing with ChatGPT, Google isn’t disclosing many technical details about how it’s training the next generation of generative AI.
Google says the new model is better at common sense, math, and logic. And it’s not just size that matters.
According to the announcement, PaLM 2 also offers enhanced code programming and debugging support. The model is trained in 20 programming languages, including popular languages such as JavaScript and Python.
Google speaks of PaLM as a family of models, including Med-PaLM 2, the business model focused on medical knowledge.
There is also Sec-PaLM, a version aimed at security applications, and a smaller PaLM-2 model that can run on smartphones.
And security?
In the future, Google wants to provide all AI-generated content with a digital watermark. And this label, which can help prevent the spread of fake news, persists even when the content is edited.
Google’s arch-rival Microsoft is known to have entered into a multibillion-dollar pact with ChatGPT developer company OpenAI, building generative AI into its applications across the board. Google has so far been reluctant to do so, citing responsible use of the technology and legal requirements.
The group confirmed this course at Google I/O. “The only way to be brave in the long run is to act responsibly from the start,” said James Manyika, who is responsible for social responsibility for the use of artificial intelligence at Google.
The group sees the danger that the software could reinforce prejudices or be used to produce and spread false information. To protect you against this, files created with the help of artificial intelligence, for example, must be provided with metadata so that they can be recognized immediately.
Google will also only provide verified developers with software that can automatically create dubbed versions of videos, Manyika said. This is to prevent so-called deepfakes with alleged actions of real people.
At the same time, Manyika emphasized that Google decided years ago not to disclose interfaces for facial recognition applications. Responsible use of artificial intelligence requires all those involved to work together.
Google Search gets AI feature
Generative AI is also being integrated into Google Search:
Similar to Microsoft’s AI-assisted Bing search, information from the web search is integrated and the websites used are visualized as links.”
The new AI search functions should initially only be available to interested testers. To try it out, you must be able to register at labs.google.com/search. However, this is not possible in Switzerland. 💩
With the help of the integrative AI, the search engine should be able to answer the question of which of the two nature parks is better suited for a family with children and a dog. The answers are formulated in full sentences, as well as the usual Internet links.
When searching for products, in addition to the previous shopping suggestions, there are AI-generated hints and tips, for example what to look for when buying a bicycle. The shopping ads would also be supplemented with AI explanations, it is said.
When searching for a vehicle, in addition to advice such as attention to the wheel suspension, suitable offers from dealers are also displayed. In addition, one could ask about hand gesture rules when cycling in California, for example, said Google executive Cathy Edwards. “These are things you would never have asked for before when searching.”
But: Google emphasizes that ‘Bard’ is an experiment and not a serious replacement for its search engine.
MusicLM – here the AI should play music
Google is also launching an experimental AI tool that can convert text to music. The “Music LM” announced in January should now be available for testing.
« Immersive View » for Google Maps
Google’s map app would get “Immersive View” for routes in some cities. A virtual preview of the route selected by the user is displayed, which among other things visualizes the current traffic as small, animated vehicles. The weather is also displayed. Regardless of whether you want to cover the route by car, bicycle or on foot.
Selected metros to receive Immersive View first include Amsterdam, Berlin, Dublin, Florence, Las Vegas, London, Los Angeles, Miami, New York, Paris, San Francisco, San Jose, Seattle, Tokyo and Venice.
A Swiss city is not among them. 💩
Why this AI offensive?
Since ChatGPT opened to the public last November, the IT industry has been busy creating generative AI systems that can generate content in seconds in response to user requests in natural language.
The public hype surrounding the capabilities of the new AI systems ranges from exuberant enthusiasm to apocalyptic fears.
Google also followed this example and made its competitor Bard available to the public to a limited extent at the end of March.
Those responsible for the American technology group never tire of emphasizing that Google is not a newcomer.
“For the past seven years, we’ve been primarily an artificial intelligence company and we’re at a turning point,” California group CEO Sundar Pichai said Wednesday night at the opening of I/O at the Google Amphitheater in Mountain View, California.
Microsoft has recently made similar announcements. The IT giant, which has invested tens of billions in Californian ChatGPT developer OpenAI, had already integrated ChatGPT into its own search engine Bing and last week fully opened AI search to the public – reviving the search portal, which is insignificant. has been compared to to Google.
What about new Google hardware?
Sources
With material from the Keystone-SDA and DPA news agencies
- blog.google: New ways AI makes Maps more immersive
- blog.google: Supercharged search with generative AI
- blog.google: Turn ideas into music with MusicLM
- golem.de: Google is bringing AI to search and apps, making Bard accessible to everyone
- theverge.com: The Pixel Fold is Google’s $1,800 introduction to the foldable phone space
- techcrunch.com: Google launches PaLM 2, its next-gen major language model
Source: Watson
I’m Ella Sammie, author specializing in the Technology sector. I have been writing for 24 Instatnt News since 2020, and am passionate about staying up to date with the latest developments in this ever-changing industry.
