Google

AI integration escalates Google’s energy use

Google’s annual sustainability report reveals a nearly 50% increase in greenhouse gas emissions from 2019 to 2024, primarily due to its data centres and supply chain. The 2024 Environmental Report indicates that Google emitted 14.3 million tons of CO2 equivalent last year, raising concerns about its goal to be net zero by 2030. The company expects emissions to rise further before declining, attributing this trend to the growing energy demands of AI integration and increased investment in technical infrastructure.

Efforts to make data centres more efficient, such as using a new generation of tensor processing units (TPUs), have been offset by the rising energy consumption required for AI. Scope 2 emissions, mainly from data centre electricity use, increased by 37% compared to 2022. The rise outpaced the company’s ability to implement carbon-free energy projects, particularly in the United States and Asia-Pacific region. Differences between Google’s global approach to carbon-free energy and the regional guidelines of the GHG Protocol have also contributed to this mismatch.

Scope 3 emissions, which account for 75% of Google’s overall emissions, rose by 8% year-on-year. These indirect emissions from the supply chain are expected to continue increasing due to capital expenditures and investments in AI-related infrastructure. A single generative AI query consumes nearly ten times the power of a regular Google search, highlighting the significant energy demands of AI technology.

Why does it matter?

Additionally, Google’s data centres consume more than three times the amount of water that Microsoft does to remain cool, underscoring the environmental challenges posed by the tech giant’s operations. The report suggests that while Google is making strides in efficiency, the rapid growth of AI and its associated infrastructure presents significant sustainability challenges.

Google warns of generative AI dangers

A recent research paper from Google reveals that generative AI already distorts socio-political reality and scientific consensus. The paper, titled ‘Generative AI Misuse: A Taxonomy of Tactics and Insights from Real-World Data,’ was co-authored by researchers from Google DeepMind, Jigsaw, and Google.org.

It categorises various ways generative AI tools are misused, analysing around 200 incidents reported in the media and academic papers between January 2023 and March 2024. Unlike warnings about hypothetical future risks, this research focuses on the real harm generative AI is currently causing, such as flooding the internet with generated text, audio, images, and videos.

The researchers found that most AI misuse involves exploiting system capabilities rather than attacking the models themselves. However, this misuse blurs the lines between authentic and deceptive content, undermining public trust. AI-generated content is being used for impersonation, creating non-consensual intimate images, and amplifying harmful content. These activities often uphold the terms of service of AI tools, highlighting a significant challenge in regulating AI misuse.

Google’s research also emphasises the environmental impact of generative AI. The increasing integration of AI into various products drives energy consumption, making it difficult to reduce emissions. Despite efforts to improve data centre efficiency, the overall rise in AI use has outpaced these gains. The paper calls for a multi-faceted approach to mitigate AI misuse, involving collaboration between policymakers, researchers, industry leaders, and civil society.

Tech giants clash over California AI legislation

California lawmakers are poised to vote on groundbreaking legislation aimed at regulating AI to prevent potential catastrophic risks, such as manipulating the state’s electric grid or aiding in the creation of chemical weapons. Spearheaded by Democratic state Sen. Scott Wiener, the bill targets AI systems with immense computing power, setting safety standards that apply only to models costing over $100 million to train.

Tech giants like Meta (Facebook) and Google strongly oppose the bill, arguing that it unfairly targets developers rather than those who misuse AI for harmful purposes. They contend that such regulations could stifle innovation and drive tech companies away from California, potentially fracturing the regulatory landscape.

While highlighting California’s role as a leader in AI adoption, Governor Gavin Newsom has not publicly endorsed the bill. His administration is concurrently exploring rules to combat AI discrimination in employment and housing, underscoring the dual challenges of promoting AI innovation while safeguarding against its misuse.

The proposed legislation has garnered support from prominent AI researchers and would establish a new state agency to oversee AI development practices and enforce compliance. Proponents argue that California must act swiftly to avoid repeating past regulatory oversights in the social media sector, despite concerns over regulatory overreach and its potential economic impact.

Google requires disclosure for election ads with altered content

Google announced that it will require advertisers to disclose election ads that use digitally altered content depicting real or realistic-looking people or events to combat misinformation during elections. This latest update to Google’s political content policy mandates advertisers to select a checkbox for ‘altered or synthetic content’ within their campaign settings.

The proliferation of generative AI, capable of rapidly creating text, images, and video, has sparked concerns over potential misuse. Deepfakes, which convincingly manipulate content to misrepresent individuals, have further blurred the distinction between fact and fiction in digital media.

To implement these changes, Google will automatically generate an in-ad disclosure for feeds and shorts on mobile devices and in-stream ads on computers and television. Advertisers must provide a prominently displayed disclosure for other ad formats that is clearly visible to users. According to Google, the exact wording of these disclosures will vary based on the context of each advertisement.

Why does it matter?

Earlier this year, during India’s general election, fake videos featuring Bollywood actors surfaced online, criticising Prime Minister Narendra Modi and urging support for the opposition Congress party. The incident highlighted the growing challenge of combating deceptive content amplified by AI-generated media.

In a related effort, OpenAI, led by Sam Altman, reported disrupting five covert influence operations in May that aimed to manipulate public opinion using AI models across various online platforms. Meta Platforms had previously committed to similar transparency measures, requiring advertisers on Facebook and Instagram to disclose the use of AI or digital tools in creating political, social, or election-related ads.

EU antitrust probe targets Microsoft-OpenAI and Google-Samsung AI deals

EU antitrust regulators scrutinise Microsoft’s partnership with OpenAI and Google’s AI deal with Samsung due to concerns over exclusivity clauses. Competition chief Margrethe Vestager plans to gather more third-party views. This development comes amid global unease about Big Tech’s dominance in new technologies.

After sending questionnaires to tech firms regarding their AI partnerships, Vestager now seeks additional information about Microsoft’s $13 billion investment in OpenAI’s for-profit subsidiary, which would result in a 49% stake, to determine if it harms competitors.

While Microsoft’s deal isn’t subject to EU merger rules, Vestager also investigates if Big Tech is blocking smaller AI developers from accessing users and businesses. Similar concerns apply to Google’s agreement to pre-install its Gemini Nano model on Samsung devices.

Vestager also examines ‘acqui-hires,’ where companies acquire others primarily for their talent, such as Microsoft’s $650-million acquisition of Inflection, to ensure these practices don’t bypass merger control rules and lead to market concentration.

Why does it matter?

Reuters reported in April that the EU regulators were building a case that could lead to an antitrust investigation into Microsoft’s $13 billion investment in OpenAI. Partnerships involving Alphabet, Amazon, and Anthropic are also under scrutiny from antitrust enforcers on both sides of the Atlantic.

AI brings 110 new languages to Google Translate

Google Translate is undergoing its largest expansion yet, adding 110 new languages and reaching nearly a tenth of the global population, thanks to advancements in AI. Google’s PaLM 2 language model has facilitated this growth by efficiently learning languages that are closely related, such as Awadhi and Marwadi, which are similar to Hindi, and French creoles like Seychellois Creole and Mauritian Creole. The expansion includes new languages like Afar, Cantonese, and Tamazight, bringing over 614 million people into the fold, with a significant focus on African languages.

Google’s approach to language variety and dialects involves creating hybrid models that prioritize the most commonly used forms. For instance, the translation model for Romani combines elements from several dialects, including Southern Vlax Romani, to produce a more universally understandable output. The strategy follows their 2022 introduction of 24 new languages using Zero-Shot Machine Translation, which enables translation without prior examples, and their ambitious 1,000 Languages Initiative to support the world’s most spoken languages.

Isaac Caswell, a Google software engineer, highlighted that ongoing partnerships with expert linguists and native speakers will further enhance the diversity and accuracy of Google Translate. The expansion underscores Google’s broader commitment to integrating AI across its platforms, a trend mirrored by competitors like Apple and Microsoft, who are also embedding AI into their services. Google Translate can be accessed at translate.google.com or through its mobile apps for Android and iOS.

The future of humour in advertising with AI

AI is revolutionising the world of advertising, particularly when it comes to humour. Traditionally, humour in advertising was heavily depended on human creativity, relying on puns, sarcasm, and funny voices to engage consumers. However, as AI advances, it is increasingly being used to create comedic content.

Neil Heymann, Global Chief Creative Officer at Accenture Song, discussed the integration of AI in humour at the Cannes Lions International Festival of Creativity. He noted that while humour in advertising carries certain risks, the potential rewards far outweigh them. Despite the challenges of maintaining a unique comedic voice in a globalised market, AI offers new opportunities for creativity and personalisation.

One notable example Heymann highlighted was a recent Uber ad in the UK featuring Robert De Niro. He emphasised that while AI might struggle to replicate the nuanced performance of an actor like De Niro, it can still be a valuable tool for generating humour. For instance, a new tool developed by Google Labs can create jokes by exploring various wordplay and puns, expanding the creative options available to writers.

Heymann believes that AI can also help navigate the complexities of global advertising. By acting as an advanced filtering system, AI can identify potential cultural pitfalls and ensure that humorous content resonates with diverse audiences without losing the thrill of creativity.

Moreover, AI’s impact on advertising extends beyond humour. Toys ‘R’ Us recently pioneered text-to-video AI-generated advertising clips, showcasing AI’s ability to revolutionise content creation across various formats. That innovation highlights the expanding role of AI in shaping the future of advertising, where technological advancements continuously redefine creative possibilities.

Meta launches AI chatbot in India, rivaling Google’s Gemini

Meta has officially introduced its AI chatbot, powered by Llama 3, to all users in India following comprehensive testing during the country’s general elections. Initially trialled on WhatsApp, Instagram, Messenger, and Facebook since April, the chatbot is now fully accessible through the search bars in these apps and the Meta.AI website. Despite supporting only English, its functionalities are on par with other major AI services like ChatGPT and Google’s Gemini, including tasks such as suggesting recipes, planning workouts, writing emails, summarising text, recommending Instagram Reels, and answering questions about Facebook posts.

The launch aims to capitalise on India’s vast user base, notably the 500 million WhatsApp users, by embedding the chatbot deeper into the user experience. However, some limitations have been observed, such as the chatbot’s inability to fully understand the context of group conversations, except in direct mentions or replies. Moreover, while it cannot be disabled, users can choose not to interact with it during searches.

Despite its capabilities, Meta AI has faced criticisms for biases in its image generation, often depicting Indian men with turbans and producing images of traditional Indian houses, which Meta has acknowledged and aims to address through ongoing updates. The launch coincides with Google releasing its Gemini app in India, which, unlike Meta’s chatbot, supports multiple local languages, potentially giving Google a competitive advantage in the linguistically diverse Indian market.

Why does it matter?

In summary, Meta’s rollout of its English-only AI chatbot in India is a strategic effort to leverage its extensive user base by offering robust functionalities similar to established competitors. While it faces initial limitations and biases, Meta is actively working on improvements. The concurrent release of Google’s Gemini app sets up a competitive landscape, underscoring the dynamic and evolving nature of AI services in India.

Google enhances Gmail with new AI features

Google is enhancing Gmail with new AI features designed to streamline email management. A new Gemini side panel is being introduced for the web, which is capable of summarising email threads and drafting new emails. Users will receive proactive prompts and can ask freeform questions, utilising Google’s advanced models like Gemini 1.5 Pro. The mobile Gmail app will also feature Gemini’s ability to summarise threads.

However, these upgrades will only be accessible to paid Gemini users. To benefit from these features, one must be a Google Workspace customer with a Gemini Business or Enterprise add-on, a Gemini Education or Education Premium subscriber, or a Google One AI Premium member. Despite their potential usefulness, it’s advised not to depend entirely on these AI tools for critical work, as AI can sometimes produce inaccurate information.

In addition to Gmail, Google is incorporating Gemini features into the side panels of Docs, Sheets, Slides, and Drive. The rollout follows Google’s earlier promises at the I/O conference. Further AI enhancements, including ‘Contextual Smart Reply,’ are expected to arrive for Gmail soon.

Brazil court ends investigation into Google, Telegram

Brazil’s top court has decided to close an investigation into Alphabet’s Google and Telegram. The tech giants were being investigated for allegedly coordinating their opposition to a bill designed to combat fake news. The pending bill would require internet companies to find and report illegal material on their platforms, with heavy fines for non-compliance.

Judge Alexandre de Moraes ruled in favour of Brazil’s deputy prosecutor general, who saw no grounds for criminal proceedings against the companies. The investigation into the executives from these tech firms, which Moraes ordered last year, will now be halted.