Google I/O 2024: Unveiling Gemini 1.5 Flash and Pro, Project Astra, and Transforming AI Search with Grounding

Mountain View, California, California, USA United States of America
An early prototype of Audio Overviews for NotebookLM was demonstrated, which creates a verbal discussion personalized for the user.
Gemini 1.5 Flash and Gemini 1.5 Pro models were announced for Google AI Studio and Vertex AI.
Gemini 1.5 Flash is a lighter-weight model designed for speed and efficiency, while Gemini 1.5 Pro offers superior general performance.
Google I/O 2024 conference featured several significant updates to its AI offerings.
Google took steps to ensure transparency and accuracy by conducting adversarial red-teaming tests on its new features.
Grounding with Google Search is now generally available on Vertex AI to connect Gemini models with world knowledge and up-to-date information.
Project Astra was unveiled as Google's vision for the future of AI assistants.
There are concerns about the potential impact of these advancements on the web economy, with some experts predicting a fall in website traffic by up to 25%.
Trillium TPU, the sixth generation custom AI accelerator, was introduced with improvements in compute performance and energy efficiency.
Google I/O 2024: Unveiling Gemini 1.5 Flash and Pro, Project Astra, and Transforming AI Search with Grounding

Google's I/O 2024 conference was a bustling hub of innovation and announcements, with several significant updates to its AI offerings taking center stage. One of the most notable announcements was the public preview release of Gemini 1.5 Flash and Gemini 1.5 Pro models on Google AI Studio and Vertex AI.

Gemini 1.5 Flash is a lighter-weight model designed for speed and efficiency, making it an ideal choice for serving at scale with a context window of up to 1 million tokens.

The Gemini 1.5 Pro model, on the other hand, offers superior general performance across a wide range of tasks. Both models are now available in public preview.

Google also unveiled Project Astra, its vision for the future of AI assistants. Additionally, Trillium TPU was introduced as the sixth generation custom AI accelerator (Tensor Processing Unit), achieving a 4.7x increase in peak compute performance per chip and over 67% more energy-efficient compared to TPU v5e.

Another exciting announcement was the demonstration of an early prototype of Audio Overviews for NotebookLM, which uses a collection of uploaded materials to create a verbal discussion personalized for the user. Grounding with Google Search, a tool that connects the Gemini model with world knowledge and up-to-date information on the internet, is now generally available on Vertex AI.

Google's I/O 2024 conference was not only about AI advancements but also addressed concerns regarding potential impacts of these technologies. For instance, audio understanding has been added to the Gemini API and AI Studio, allowing Gemini 1.5 Pro to reason across image and audio for videos uploaded in AI Studio.

Google took steps to ensure transparency and accuracy by conducting adversarial red-teaming tests on its new features. The company expects that Project Astra's implementation of generative AI will be rooted in Search's core quality and safety systems, with built-in guardrails to prevent low-quality or harmful information from surfacing.

Despite these advancements, there are concerns about the potential impact on the web economy. Some experts predict that traffic to websites could fall by up to 25% due to increased reliance on AI search results and summaries. This could result in significant revenue loss for publishers and content creators.

Google maintains that its new formula will encourage users to click on a wider variety of websites, not less. News outlets and other media creators rich with fresh information could potentially strike deals with Google to license their data used in AI models. However, the full implications of these changes remain to be seen.


Article Image
Article Image
Article Image
Article Image
Article Image
Article Image
Article Image
Article Image
Article Image
Article Image

Confidence

90%

Doubts
  • Are there any potential drawbacks to the increased reliance on AI search results?
  • How will publishers and content creators be affected by these changes in web traffic patterns?
  • Will Google's guardrails be effective in preventing low-quality or harmful information from surfacing?

Sources

77%

The overall score is a weighted number that takes into account conflict of interest, bias, deception and other practices that undermine the credibility of the source. It is calculated as:
(Site Conflicts Of Interest + Author Conflicts Of Interest) / 2.0 * 0.2 + ArticleBiasScore * 0.20 + UniquePointsScore * 0.05 + DeceptionScore * 0.20 + ReadabilityScore * 0.05 + FallacyScore * 0.20

A score that takes into consideration the content for flow, interruptions with ads, and overt search engine optimization techniques that makes the content hard to understand

Publishers Horrified at New Google AI Feature That Could Kill What's Left of Journalism

Futurism News Site Description and Analysis Report 2021 Update Sharon Adarlo Friday, 17 May 2024 02:13
  • Unique Points
    • Google has introduced a new AI feature called AI Overviews in its search engine.
    • News publishers are concerned that this will reduce traffic to their websites and potentially lead to significant revenue loss.
    • Preliminary studies suggest that website traffic could be reduced by up to 25% as a result of Google’s use of AI in its search engine.
  • Accuracy
    • ]Google has introduced a new AI feature called AI Overviews in its search engine.[
  • Deception (50%)
    The author uses emotional manipulation by describing the potential impact of Google's new AI feature as an 'extinction-level event' and a 'death knell for any remaining traffic to media websites'. She also engages in selective reporting by only mentioning the potential negative consequences of the feature without providing any counterarguments or evidence that other benefits may exist. The author also uses sensational language, such as 'this has the potential to be an extinction-level event' and 'billions in revenue lost'.
    • The little traffic we get today will be further diminished...
    • This new feature, powered by Google’s Gemini AI model, will summarize answers to search queries as the top result for many queries...
    • This will be catastrophic to our traffic...
  • Fallacies (85%)
    The author uses an inflammatory and hyperbolic comparison of Google's new AI feature to an asteroid impact that killed the dinosaurs, implying a catastrophic effect on journalism traffic. This is an example of a fallacy called Hyperbole.
    • This will be catastrophic to our traffic, as marketed by Google to further satisfy user queries, leaving even less incentive to click through so that we can monetize our content.
    • This has the potential to be an extinction-level event akin to when an asteroid slammed into Mexico 66 million years ago, ending the reign of the journalism dinosaurs.
  • Bias (80%)
    The author expresses a clear bias against Google's new AI feature and its potential impact on news publishers. She uses strong language such as 'catastrophic', 'extinction-level event', and 'perverse twist' to describe the situation.
    • "This time with a product that directly competes with our content, using our content to fuel it. This is a perverse twist on Ňinnovation.ň
      • "This will be catastrophic to our traffic,"
      • Site Conflicts Of Interest (100%)
        None Found At Time Of Publication
      • Author Conflicts Of Interest (100%)
        None Found At Time Of Publication

      98%

      The overall score is a weighted number that takes into account conflict of interest, bias, deception and other practices that undermine the credibility of the source. It is calculated as:
      (Site Conflicts Of Interest + Author Conflicts Of Interest) / 2.0 * 0.2 + ArticleBiasScore * 0.20 + UniquePointsScore * 0.05 + DeceptionScore * 0.20 + ReadabilityScore * 0.05 + FallacyScore * 0.20

      A score that takes into consideration the content for flow, interruptions with ads, and overt search engine optimization techniques that makes the content hard to understand

      100 things we announced at I/O 2024

      The Keyword by Google Molly McHugh-Johnson Wednesday, 15 May 2024 00:00
      • Unique Points
        • Gemini 1.5 Flash, a lighter-weight model designed for speed and efficiency, is now available in public preview with a 1 million token context window on Google AI Studio and Vertex AI.
        • Gemini 1.5 Pro, the best model for general performance across a wide range of tasks, is also available in public preview with a 2 million token context window to developers via waitlist in Google AI Studio and Vertex AI.
        • Project Astra, Google’s vision for the future of AI assistants, was announced.
        • Trillium TPU, the sixth generation of Google’s custom AI accelerator, is now available. It achieves a 4.7x increase in peak compute performance per chip and is over 67% more energy-efficient compared to TPU v5e.
        • An early prototype of Audio Overviews for NotebookLM was demonstrated, which uses a collection of uploaded materials to create a verbal discussion personalized for the user.
        • Grounding with Google Search, a tool that connects the Gemini model with world knowledge and up-to-date information on the internet, is now generally available on Vertex AI.
        • Audio understanding has been added to the Gemini API and AI Studio, allowing Gemini 1.5 Pro to reason across image and audio for videos uploaded in AI Studio.
      • Accuracy
        No Contradictions at Time Of Publication
      • Deception (100%)
        None Found At Time Of Publication
      • Fallacies (95%)
        The article contains some inflammatory rhetoric and appeals to authority. It also uses a dichotomous depiction by presenting Google's AI tools as superior to other foundation models without providing evidence or comparison data.
        • . . . whether you were most into the latest Gemini app updates, felt especially excited about what’s coming for developers or can’t wait to try the latest generative AI tools, there was something for just about everyone.
        • We introduced Gemini 1.5 Flash: a lighter-weight model that’s designed to be fast and efficient to serve at scale. . . We’ve significantly improved 1.5 Pro, our best model for general performance across a wide range of tasks.
        • We shared Project Astra: our vision for the future of AI assistants.
        • And we demoed an early prototype of Audio Overviews for NotebookLM, which uses a collection of uploaded materials to create a verbal discussion personalized for the user.
      • Bias (100%)
        None Found At Time Of Publication
      • Site Conflicts Of Interest (100%)
        None Found At Time Of Publication
      • Author Conflicts Of Interest (100%)
        None Found At Time Of Publication

      94%

      The overall score is a weighted number that takes into account conflict of interest, bias, deception and other practices that undermine the credibility of the source. It is calculated as:
      (Site Conflicts Of Interest + Author Conflicts Of Interest) / 2.0 * 0.2 + ArticleBiasScore * 0.20 + UniquePointsScore * 0.05 + DeceptionScore * 0.20 + ReadabilityScore * 0.05 + FallacyScore * 0.20

      A score that takes into consideration the content for flow, interruptions with ads, and overt search engine optimization techniques that makes the content hard to understand

      How Google’s AI Overviews Work, and How to Turn Them Off (You Can’t)

      Wired Magazine Reece Rogers Thursday, 16 May 2024 21:42
      • Unique Points
        • Google’s AI Overviews appear for complex queries.
        • AI Overviews are designed to highlight information that can be easily verified by the supporting information Google surfaces.
        • For some AI Overviews, webpage links are immediately visible, for others you have to click ‘Show more.’
        • Google did adversarial red-teaming tests to uncover potential weak points for the feature.
        • Google’s Liz Reid expects AI Overviews to arrive for countries outside of the United States before the end of 2024.
      • Accuracy
        • Google's AI Overviews appear for complex queries.
        • The new feature summarizes answers to search queries as the top result, with links appearing below.
        • Google executes adversarial red-teaming tests to uncover potential weak points for the feature.
        • Google's Liz Reid expects AI Overviews to arrive for countries outside of the United States before the end of 2024.
        • Google is introducing AI-generated answers to online queries in the United States, marking one of the biggest changes to its world-leading search engine in 25 years.
        • The change will soon spread to other countries, potentially reducing the importance of links and web pages for over a billion people.
      • Deception (100%)
        None Found At Time Of Publication
      • Fallacies (85%)
        The author makes an appeal to authority when quoting Mallory De Leon and Liz Reid. They also use inflammatory rhetoric by expressing their apprehension about the AI Overviews and questioning their trustworthiness.
        • “AI Overviews appear for complex queries,” says Mallory De Leon, a Google spokesperson.
        • “This implementation of generative AI is rooted in Search’s core quality and safety systems,” she says.
        • “As someone whose job relies on readers actually clicking links and spending time reading the articles, of course I’m apprehensive about this change⁄
        • “Even though AI Overviews are designed to save you time, they might lead to less trustworthy results.
      • Bias (95%)
        The author expresses concern about the potential for inaccuracies or 'lies' in Google's AI-generated summaries (AI Overviews) and mentions that there is no disclaimer warning users of this possibility. This demonstrates a bias against the use of AI-generated content, implying that it may be unreliable or even harmful.
        • There's no such disclaimer added to the bottom of the AI Overview, which often simply reads, ‘Generative AI is experimental.’
          • Yes, OpenAI’s ChatGPT and other AI tools are quite popular in Silicon Valley tech circles, but this feature will likely expose billions of people, who have never used a chatbot before, to AI-generated text. Even though AI Overviews are designed to save you time, they might lead to less trustworthy results.
          • Site Conflicts Of Interest (100%)
            None Found At Time Of Publication
          • Author Conflicts Of Interest (100%)
            None Found At Time Of Publication

          88%

          The overall score is a weighted number that takes into account conflict of interest, bias, deception and other practices that undermine the credibility of the source. It is calculated as:
          (Site Conflicts Of Interest + Author Conflicts Of Interest) / 2.0 * 0.2 + ArticleBiasScore * 0.20 + UniquePointsScore * 0.05 + DeceptionScore * 0.20 + ReadabilityScore * 0.05 + FallacyScore * 0.20

          A score that takes into consideration the content for flow, interruptions with ads, and overt search engine optimization techniques that makes the content hard to understand

          Google's AI search revamp puts publishers in a quandary

          Entertainment Variety Tv Site: https://www.yahoo.com/entertainment/tv-site/ Glenn CHAPMAN Friday, 17 May 2024 02:17
          • Unique Points
            • Google is introducing AI-generated answers to online queries in the United States, marking one of the biggest changes to its world-leading search engine in 25 years.
            • Google executives maintain that the new formula would encourage users to click on a wider variety of websites, not less.
            • News outlets and other media creators rich with fresh information could potentially strike deals with Google to make money from licensing the data used in AI models.
            • Research firm Gartner predicts traffic to the web from search engines will fall 25% by 2026 due to increased reliance on AI.
            • The advertising industry may face a significant hit of billions of dollars as a result of these changes.
          • Accuracy
            • The change will soon spread to other countries, potentially reducing the importance of links and web pages for over a billion people.
            • Research firm Gartner predicts traffic to the web from search engines will fall 25% by 2026 due to increased reliance on AI.
            • News outlets and other media creators rich with fresh information could potentially strike deals with Google to make money from licensing the data used in AI models.
          • Deception (70%)
            The article contains editorializing and selective reporting. The author expresses opinions about the potential impact of Google's AI search revamp on publishers and traffic to their websites. He quotes Paul Roetzer stating that 'It’s going to create a negative impact on brands and publishers who rely on organic search traffic for sure.' This is an opinion, not a fact. The author also mentions research from Gartner predicting a 25% decrease in web traffic from search engines by 2026 without providing any context or citation for this claim. This selective reporting of information supports the author's opinion and creates a sensationalist tone.
            • Marketers and publishers have little choice but to keep doing what they are doing, and diversify where they appear online to get noticed in places other than Google searches.
            • Roetzer noted that Google has not provided much information about how the change might affect advertisers or publishers, essentially asking them to have faith.
            • The change will soon spread to other countries, arguably reducing the importance of links and web pages for more than a billion people.
          • Fallacies (85%)
            The article contains a few informal fallacies and appeals to authority. It also presents dichotomous depictions of Google's AI search revamp.
            • . . . publishers wondering if traffic to their websites will wither.
            • Google's use of artificial intelligence to sum up answers to search queries has publishers wondering if traffic to their websites will wither.
            • For now, marketers and publishers have little choice but to keep doing what they are doing, and diversify where they appear online to get noticed in places other than Google searches.
          • Bias (95%)
            The author expresses concern for publishers and their potential loss of traffic due to Google's new AI-generated answers. The author quotes Paul Roetzer stating that the change will have a negative impact on brands and publishers, and that they don't know what to do about it. The author also mentions research predicting a 25% decrease in web traffic from search engines by 2026 due to increased reliance on AI. These statements demonstrate a bias towards the perspective of publishers and their potential loss.
            • Google’s use of artificial intelligence to sum up answers to search queries has publishers wondering if traffic to their websites will wither.
              • It’s going to create a negative impact on brands and publishers who rely on organic search traffic for sure.
                • We just have no idea how much, and we don’t really know what you can do about it.
                • Site Conflicts Of Interest (100%)
                  None Found At Time Of Publication
                • Author Conflicts Of Interest (100%)
                  None Found At Time Of Publication

                97%

                The overall score is a weighted number that takes into account conflict of interest, bias, deception and other practices that undermine the credibility of the source. It is calculated as:
                (Site Conflicts Of Interest + Author Conflicts Of Interest) / 2.0 * 0.2 + ArticleBiasScore * 0.20 + UniquePointsScore * 0.05 + DeceptionScore * 0.20 + ReadabilityScore * 0.05 + FallacyScore * 0.20

                A score that takes into consideration the content for flow, interruptions with ads, and overt search engine optimization techniques that makes the content hard to understand

                Google’s New Search Engine AI Overviews Is Bad News for the Web Economy

                Bloomberg News Now Dave Lee Friday, 17 May 2024 02:18
                • Unique Points
                  • Google unveiled 'AI Overviews' at its annual developers conference.
                  • It is a significant update to Google's search engine.
                • Accuracy
                  • If successful, it could reduce the amount of traffic websites receive.
                  • Preliminary studies suggest that website traffic could be reduced by up to 25% as a result of Google’s use of AI in its search engine.
                • Deception (100%)
                  None Found At Time Of Publication
                • Fallacies (100%)
                  None Found At Time Of Publication
                • Bias (100%)
                  None Found At Time Of Publication
                • Site Conflicts Of Interest (100%)
                  None Found At Time Of Publication
                • Author Conflicts Of Interest (100%)
                  None Found At Time Of Publication