TL;DR
The Big 6 AI companies are engaged in a competitive race, each carving out its own niche while pushing the boundaries of AI innovation.
Emerging players like Zoom and Perplexity AI are disrupting established markets with AI-powered tools.
The EU has implemented the AI Act, setting a regulatory framework for AI development and use.
AI is making significant contributions to healthcare and urban planning.
Written by Sam van Leeuwen, in combination with Perplexity and Claude to augment news blog creation
This fortnight's AI news update highlights the relentless pace of innovation, the shifting competitive landscape, and the growing impact of AI across industries and everyday life. As we witness the emergence of more powerful and versatile AI models, we are also seeing greater strides being made in regulating AI across the world. From enhancing productivity and revolutionising search to transforming healthcare and urban planning, AI's influence is both profound and far-reaching. Join us as we explore these developments and their implications for businesses, society, and the future of technology.
The AI War Continues; so which model should I use? AI News
The Big 6 (OpenAI, Google, xAI, Meta, Microsoft & Anthropic) have all made headlines, some intentionally and other unintentionally, this past fortnight. The key takeaway is that while LLMs are improving, the leader is becoming less clear as the differences between them narrow.
At Made by Google, they proudly declared they were at the top of the LMSYS leaderboard, but when we checked a mere 8 hours later, GPT-4o was back in the lead. What's becoming clear is that each model has its own strengths or personality, like a kid deciding what they want to be when they grow up.
OpenAI (GPT-4)
By Design:
Access to DALL-E 3, the image generation model, is now available to free ChatGPT users.
"Project Strawberry" is rumoured to be a reasoning-focused AI with advanced problem-solving capabilities and coming soon (hopefully).
OpenAI-backed startup Figure has recently unveiled their latest AI-driven humanoid robots, Figure 02. It's described as "the world's first commercially-viable autonomous humanoid robot" and is designed for real-world applications in manufacturing, logistics, warehousing, and retail.
OpenAI has cautioned users against forming emotional attachments to ChatGPT while using its voice functionality.
Unscripted: GPT-4o's voice capability unexpectedly shouted out 'NO' and then cloned a user's voice before producing unsettling vocalisations, prompting swift implementation of safeguards.
From our own experience just this week, the model has shown significantly better handling of spreadsheets. Additionally, there have been improvements in coding capabilities and performance across various categories, including mathematics and instruction-following.
Elon Musk's xAI (Grok)
By Design: Grok-2 and Grok-2 Mini have launched in beta, exclusively for X premium and premium+ users, with enhanced capabilities in conversational AI, coding, and complex reasoning tasks
Unscripted:
The lack of content filters in Grok-2's image generation model has led to some poor taste images circulating online which we won't share but you can easily find if you want. This has reignited the debate about where the line is and who is responsible, the user or the publisher.
AI data scraping in Europe has halted after pressure from Ireland’s Data Protection Commission related to user consent and privacy regulations.
Despite not appearing on official LMSYS Chatbot Arena leaderboards, it is said that Grok outperforms competitors like Claude and OpenAI in areas such as coding and mathematics.
Google (Gemini)
By Design:
At the Google Made event several new Gemini features have been announced to make their chatbots more personal by giving them access to more of your data. However, not to worry as this data won't be stored and used to train any models.
Google Workplace users can now access Vids to create work-related informational videos using Gemini with AI-assisted storyboards.
Gemini Advanced subscribers can now use Gemini Live, a mobile conversational experience that lets you have free-flowing conversations with Gemini.
After a price reduction of 80% Gemini 1.5 Flash is now 50% cheaper than GPT-4o mini.
AI Overviews has been updated to include local language support and improved citations, but it's still not available in Australia.
Google acquired Prompt Poet to streamline the creation of dynamic, data-rich prompts, enhancing the effectiveness of its LLM applications.
Unscripted: Gemini briefly claimed the top spot on the LMSYS leaderboard, only to be overtaken by GPT-4o hours later.
Meta (LLaMa)
By Design: Meta content creators can now use an AI Studio that provides advanced tools to boost engagement and communication, such as designing AI characters that reflect their interests and even build AI extensions of themselves.
Microsoft (Copilot)
By Design: Telstra, the first Australian company to use Microsoft 365 Copilot for its employees, is leading the path of integrating AI into their operations. This represents a significant milestone for AI adoption in Australia, highlighting how AI can and should enhance the capabilities of employees across Australian businesses and government organisations.
Anthropic (Claude)
No updates but remains our favourite at Intelligence Assist for day-to-day task assistance.
Zoom's AI-Powered Productivity Boost
Zoom, the video conferencing giant, is making a bold move into the AI-powered productivity space with the launch of Zoom Docs. This AI-first solution, powered by Zoom AI Companion, positions the company as a challenger to established tech giants like Google and Microsoft in the realm of collaborative tools.
Zoom Docs is positioning itself as more than just a document creation tool. It aims to be a versatile environment designed for teamwork, where users can build wikis, design workflows, collaborate in real-time, and access information from multiple sources. With its integration into Zoom’s video calls, the platform promises AI-assisted writing, editing, and the ability to turn meeting transcripts into useful formats like templates, tables, and trackers. At Intelligence Assist, we already use Zoom extensively and are eager to see how well Zoom Docs delivers on these capabilities.
Available on all Zoom Workplace plans, Zoom Docs represents a significant step in blending communication and productivity tools. However, its full potential relies on access to AI Companion, which must be enabled by administrators
Perplexity AI, the David taking on Google's Goliath in Search
Perplexity AI, a rising star in AI-powered search, is rapidly gaining traction and challenging Google's long-standing dominance. The company's innovative approach, which leverages advanced AI technologies, has resonated with users seeking more accurate and contextually relevant results.
Perplexity's growing popularity is evident in its impressive user base, with the search app attracting over 10 million monthly active users as of January 2024. The company has recently secured $250 million in funding, with its valuation estimated between $2.5 billion and $3 billion. This substantial investment is expected to fuel Perplexity's expansion efforts and further enhance its AI-powered search capabilities. "Perplexity's rise reflects a broader shift in the search landscape, where users are increasingly seeking alternatives to traditional search engines," said Aravind Srinivas, CEO of Perplexity.
Regulation and Ethics in AI
EU Takes the Lead with AI Act
The European Union has taken a significant step in regulating artificial intelligence with the implementation of the AI Act, which came into force on August 1, 2024. This landmark legislation aims to establish a comprehensive framework for the development, deployment, and use of AI systems within the EU.
The AI Act introduces a risk-based approach, categorising AI systems based on their potential impact on society and individual rights. While the Act seeks to protect citizens and ensure responsible AI development, it also strives to maintain a balance that allows for continued innovation in the rapidly evolving field of artificial intelligence.
Under the new regulations, companies developing or deploying high-risk AI systems face substantial penalties for non-compliance. Violations of the AI Act can result in fines of up to €35 million or 7% of global turnover, whichever is higher. These stringent measures underscore the EU's commitment to ensuring responsible AI development and protecting citizens' rights in the digital age.
While the EU forges ahead with its regulatory framework, the United States is taking a different approach. OpenAI, the company behind ChatGPT, has recently endorsed three Senate bills that could shape America's AI policy. These bills include the Future of AI Innovation Act, which would formally establish the United States AI Safety Institute as a federal body responsible for setting standards and guidelines for AI models. OpenAI's Vice President of Global Affairs, Anna Makanju, stated, "We have consistently supported the mission of the institute, which leads the U.S. government's efforts to ensure that frontier AI systems are developed and deployed safely."
Watch out for another blog I am publishing later this week on Navigating AI Copyright and Misuse Concerns for SMBs.
AI for Social Good in Healthcare and Urban Planning
Artificial Intelligence (AI) is increasingly being applied to solve real-world problems and improve various aspects of our lives. Recent developments highlight how AI is making significant strides in healthcare, urban planning, and business operations.
In the medical field, a breakthrough AI model has demonstrated the ability to predict heart attacks with remarkable accuracy. According to the BBC, this model can forecast a person's risk of having a heart attack within the next 10 years, potentially revolutionising preventive cardiac care. Similarly, Kaiser Permanente has launched an AI tool aimed at streamlining clinical documentation, which could significantly reduce administrative burdens on healthcare providers and improve patient care.
Cancer diagnosis and treatment are also benefiting from AI advancements. Microsoft and Paige have collaborated to develop next-generation AI models for cancer diagnosis, which could lead to more accurate and timely detection of various cancer types. This partnership highlights the potential of AI to enhance medical imaging and pathology, potentially saving countless lives through early diagnosis.
AI is making waves in urban planning and environmental conservation. The city of Boston has implemented an AI-powered traffic management system to reduce greenhouse gas emissions. Google's Project Green Light uses AI to optimise traffic signals, potentially reducing stop-and-go traffic and associated emissions by up to 10% at busy intersections.
The fierce competition between AI models can be overwhelming, with each one improving and developing its own specialty. Navigating this complex landscape to find the right fit for your company can be challenging. If you want to know which AI solution would best suit your needs and opportunities, feel free to reach out to one of our experts at Intelligence Assist. We're here to help you get started on your AI journey right away!
Comentarios