OpenAI's Search Bombshell: Google's AI Reckoning Begins
10 May 2024:OpenAI takes on Google with AI search, Model Spec guides AI behaviour, businesses lack AI training, TikTok labels AI content, & AI investments pour in.
Good Morning & Happy Friday,
Latest in AI,
1. 1-Minute Roundup
2. Headlines
3. AI Investment Deals
4. Other AI News
1-Minute Roundup
Headlines:
OpenAI will launch an AI-powered search product on Monday, challenging Google's dominance.
ISACA study finds widespread AI use in businesses despite limited employee training.
OpenAI releases Model Spec, a framework guiding AI tool responses to prompts.
AI Investment Deals:
Dendra Systems secures $15.76M for an AI-based ecological restoration platform.
Hexis raises $2M for an AI-powered sports nutrition app.
Greenboard raises $4.5M for an AI-driven financial compliance platform.
In-House Health secures $4M for AI-enabled nursing scheduling platform.
RunPod raises $20M for AI application deployment and scaling platform.
Archive Intel raises $1M for AI-driven communications archiving and compliance.
Fairgen raises $8M for generative AI models optimising cancer centre workflows.
Sift Healthcare secures $20M for AI-powered healthcare payment solutions.
Yellow raises $5M for Gen AI-powered 3D modelling tool.
Others:
OpenAI joins C2PA to promote digital content transparency.
Google I/O 2024 will focus on the generative AI platform Gemini.
Retell AI offers a platform for creating AI-powered voice agents.
SoftBank is in talks to acquire struggling AI chip startup Graphcore.
TikTok to auto-label AI-generated content from platforms like DALL·E 3.
Headlines
OpenAI to Launch Google Search Rival on Monday
OpenAI plans to unveil its AI-powered search product, an extension of ChatGPT, on Monday, potentially challenging Google's search dominance and competing with AI search startup Perplexity.
Why it Matters: OpenAI's entry into the search market with an AI-powered product could disrupt the industry, putting pressure on Google and other players to innovate and improve their offerings.
Zoom In:
The announcement will be made a day before Google's annual I/O conference.
OpenAI's search product will enable ChatGPT to pull information directly from the web and include citations.
The product addresses ChatGPT's struggles with providing accurate and real-time information.
Context: AI companies focus on the search industry, and many are developing chatbots and other AI-powered tools to compete in the market. OpenAI's ChatGPT has already been integrated with Microsoft's Bing search engine.
What Others Say: Industry observers have long considered ChatGPT an alternative for gathering online information despite its challenges in providing accurate, real-time data from the web.
Between the Lines: The timing of OpenAI's announcement to launch the following Monday, just before Google's I/O conference on Tuesday, suggests a strategic move to capture attention and challenge the search giant's dominance in the market.
Zoom Out: OpenAI's search product launch is part of a broader trend of AI companies expanding their offerings and competing in the search market, putting pressure on established players like Google to innovate and adapt.
What to Watch: OpenAI's search product's success will depend on its ability to provide accurate, real-time information and a user-friendly interface that can compete with Google and other AI-powered search tools.
What's Next: As OpenAI enters the search market, the industry can expect increased competition and innovation, with companies vying to develop the most advanced and user-friendly AI-powered search products.
AI Adoption Outpaces Training in Businesses, ISACA Finds
ISACA's study highlights the widespread use of AI in businesses despite limited employee training and formal AI policies, emphasising the urgent need for upskilling to protect organisations and harness AI's benefits effectively.
Why it Matters: The rapid adoption of AI without adequate training and policies leaves businesses vulnerable to data breaches and internal skills gaps, underscoring the importance of developing a clear AI strategy and investing in employee education.
Zoom In:
74% of businesses say their staff use AI at work.
Only 30% of European organisations provide limited AI training to tech-related employees.
17% of organisations have formal, comprehensive AI policies in place.
70% of businesses offer little to no AI training for employees.
40% of European organisations provide no AI training at all.
74% of respondents are only somewhat familiar or unfamiliar with AI.
Between the Lines: Despite the technology's widespread use, businesses need AI training and policies. This suggests a disconnect between adoption and preparedness, potentially exposing organisations to risks.
Big Picture: As AI continues to shape the IT and cybersecurity industries, businesses must prioritise upskilling and developing comprehensive AI strategies to stay competitive and secure.
What to Watch:
61% of respondents are highly apprehensive about Gen AI being exploited by bad actors.
86% of respondents believe AI training will be necessary to retain their jobs or advance their careers within the next two years.
OpenAI Proposes Model Spec for AI Behavior
OpenAI releases the first draft of Model Spec, which guides how AI tools like GPT-4 respond to prompts. The framework addresses information hazards, NSFW content, and legal compliance.
Why it Matters: Establishing clear guidelines for AI model behaviour is crucial for addressing issues like misinformation, hate speech, and potential misuse while fostering trust and responsible development of AI technologies.
Zoom In:
Model Spec proposes three general principles: assist developers and users, benefit humanity, and reflect well on OpenAI.
The framework includes rules such as following the chain of command, complying with laws, and protecting privacy.
Context: Recent incidents of AI tools behaving unexpectedly, such as Microsoft's Bing AI losing track of time or Google's Gemini image generator drawing diverse Nazis, have highlighted the need for more precise guidelines on AI model behaviour.
How it Works: OpenAI's Model Spec suggests default behaviours for AI models, such as assuming users' best intentions, asking clarifying questions, taking an objective point of view, and discouraging hate.
Between the Lines: The release of Model Spec as a living document open to public feedback suggests OpenAI's desire to engage with various stakeholders in shaping the future of AI model behaviour, although the company ultimately has the final say on its models.
Reality Check: Model Spec will not immediately impact OpenAI's currently released models, which continue to operate under their existing usage policies. The company still needs to provide a timeframe for incorporating public feedback into future drafts.
What's Next: OpenAI will gather feedback from the public and various stakeholders to inform future iterations of Model Spec and responsibly build towards its mission of developing safe and beneficial AI.
AI Investment Deals
Dendra Systems secures $15.76M in Series B funding, led by Zouk Capital, to expand its AI-based ecological restoration platform globally.
Hexis, an AI-powered sports nutrition app, raises $2M in pre-seed funding, co-led by Apex Capital and Sports Republic, to expand its technology and marketing.
Greenboard raises $4.5M in seed funding to develop an AI-powered platform for financial compliance and back office management, led by Base 10 Partners.
In-House Health raises $4M in seed funding led by NEA and TMV to develop its AI-enabled scheduling platform for nursing teams, aiming to reduce labour costs and improve efficiency.
RunPod raises $20M in Seed funding, co-led by Intel Capital and Dell Technologies Capital, to enhance its cloud platform for AI application deployment and scaling.
Archive Intel raises $1M in Seed funding led by Social Leverage to enhance its AI-driven communications archiving and compliance solutions.
Fairgen, based in Tel Aviv, raises $8M in Seed funding to enhance its AI technology for generating survey responses, led by Maverick Ventures Israel.
Triomics raises $15M to expand its generative AI models to optimise workflows in cancer centres, with backing from Lightspeed, Nexus Venture Partners, General Catalyst, and Y Combinator.
Sift Healthcare secures $20M in Series B funding led by B Capital to expand its AI-powered healthcare payment solutions and Payments Intelligence Platform.
Yellow raises $5M from A16z Games to develop its Gen AI-powered 3D modelling tool, YellowSculpt, which simplifies the creation of pre-rigged 3D character meshes from text prompts.
Others
OpenAI Joins C2PA Steering Committee to Promote Digital Content Transparency
OpenAI will collaborate with C2PA members to develop and promote the adoption of Content Credentials.
OpenAI began attaching Content Credentials to DALLE 3 images and plans to do so for Sora videos.
Google I/O 2024 Preview: AI to Take Center Stage
Generative AI platform Gemini is expected to be the main focus, with deeper integration into Google services.
Android 15 developer preview may offer a glimpse into new features and Gemini (AI) integration.
Retell AI Offers Platform for Creating AI-Powered Voice Agents
Retell AI raised $4.53M to develop a platform for building AI voice agents for customer service.
The global contact centre AI market could reach nearly $3B by 2028, up from $2.4B in 2022.
SoftBank in Advanced Talks to Acquire Struggling AI Chip Startup Graphcore
SoftBank and Graphcore have been in discussions for months, with talks recently becoming more advanced.
Graphcore, once valued at $2.8B, reported a 46% revenue drop and widening losses in 2022.
TikTok to Auto-Label AI-Generated Content from Platforms like DALL·E 3
TikTok implements Content Credentials to label AI-generated content from other platforms automatically.
Change rolls out globally, with TikTok AI effects to include Content Credentials in coming months.
Like TheAIML?
Please don’t keep it to yourself. Colleagues and friends can sign up here.
For even more,
Our newsletter archive: Click here
Follow us on Linkedin,
Follow us on 𝕏/Twitter
We value your opinion. Please take a moment to share your thoughts on this newsletter. Your feedback is crucial in helping us improve and deliver the content you want to see. Please Email us