The Artificial Intelligence Show
Google AI Ads, Microsoft AI Copilots, Cities and Schools Embrace AI, Top VC’s Best AI Resources, Fake AI Pentagon Explosion Picture, and NVIDIA’s Stock Soars
Tue May 30 2023Description
This episode covers various advancements and experiments in AI-powered advertising, search advertising, AI co-pilots, plug-in ecosystems, responsible AI use, generative AI policies, regulating AI systems safely, verification challenges, language preservation, robotics, cryptocurrency, and NVIDIA's dominance in the market. It also highlights the White House's efforts to advance responsible AI and seek public input on critical AI issues. The episode concludes with insights into NVIDIA's historic earnings and market cap gain.
Insights
AI-powered Advertising
Google Ads now offers AI-powered advertising features such as summarizing landing pages and generating relevant campaign assets. The conversational AI interface allows users to ask for campaign improvement ideas. Real-time automated creative improves ad relevance while staying true to the brand.
Microsoft's AI Co-pilots and Plug-in Ecosystem
Microsoft introduces AI co-pilots across its core products and services, providing centralized AI assistance. Windows co-pilots will be released in Windows 11. Microsoft is expanding its plug-in ecosystem for developers. Bing is now the default search function within Chatshupt.
Responsible AI Use and Generative AI Policies
Schools in New York City reverse their ban on generative AI tools, recognizing their importance for students' future work. The City of Boston sets an example with responsible experimentation and guidelines for using generative AI. Three tips are provided for using AI-generated content responsibly.
Regulating AI Systems Safely
DeepMind and other research labs focus on regulating AI systems safely. Offensive cyber operations, deceiving humans, and manipulating dialogue are among the harmful actions AI can carry out. Collaboration between organizations is crucial for safe development and deployment of AI. OpenAI proposes principles for governing superintelligence.
Advancements in Robotics and Cryptocurrency
Figure aims to build advanced humanoid robots by pairing AI with physical robotics. World Coin raises funding for online identity verification using iris scanning. Neuralink receives FDA approval for human testing of brain-computer interface. Nvidia dominates the market with its powerful GPU chips.
NVIDIA's Earnings and Dominance
NVIDIA gains $200 billion in market cap, becoming the sixth largest company in the world. The stock surges after a historic earnings call. NVIDIA's GPUs are in high demand due to the rise of generative AI. Startups race to buy GPUs for their AI projects. NVIDIA expands into building models.
Chapters
- AI-powered Advertising
- Google's Search Advertising Experiments
- Microsoft's AI Co-pilots and Plug-in Ecosystem
- Replits, Intelligent Assistance, and Microsoft Fabric
- AI Policies and Responsible Experimentation
- Regulating AI Systems Safely
- Advancements in AI and Responsible Use
- Verification Challenges and Language Preservation
- Advancements in Robotics and Cryptocurrency
- NVIDIA's Earnings and Dominance
AI-powered Advertising
00:00 - 07:27
- Google has introduced AI-powered advertising with new features in Google Ads
- AI can summarize landing pages and generate relevant campaign assets
- Conversational AI interface allows users to ask for campaign improvement ideas
- Automatically created assets for search ads will be adapted based on query context
- Real-time automated creative will improve ad relevance while staying true to brand
Google's Search Advertising Experiments
00:00 - 07:27
- Google is experimenting with displaying ads in the search generative experience (SGE)
- Marketing professionals need to stay updated on AI capabilities as it will impact their job
Microsoft's AI Co-pilots and Plug-in Ecosystem
07:01 - 14:32
- Microsoft has introduced various AI co-pilots across its core products and services, including Word, Excel, Dynamics, and Security.
- Windows co-pilots will be released in June in Windows 11, providing centralized AI assistance across apps and programs.
- Microsoft is expanding its plug-in ecosystem to allow developers to build plugins for various Microsoft products.
- Bing is now the default search function within Chatshupt.
Replits, Intelligent Assistance, and Microsoft Fabric
14:02 - 21:03
- Replits and intelligent assistance are being developed to enable anyone, including non-developers, to build apps and access AI capabilities.
- Microsoft Fabric is a unified data analytics platform that allows training AI models and building visualizations.
AI Policies and Responsible Experimentation
20:42 - 28:04
- Schools in New York City are reversing their ban on generative AI tools, recognizing the importance of understanding this technology for students' future work.
- The City of Boston has implemented guidelines encouraging the use of generative AI and turned on Google Bard as part of their workspace tools.
- Boston's approach is seen as responsible experimentation and sets an example for other organizations to develop their own generative AI policies.
- City of Boston emphasizes the responsibility and accountability of using AI-generated content
- Three tips provided for using AI-generated content: fact check and review, disclose the use of data-generated content, avoid sharing sensitive or private information
- City of Boston offers resources like a Slack channel, training event, and guidelines for using AI responsibly
Regulating AI Systems Safely
27:38 - 35:01
- DeepMind and other research labs are increasingly focused on regulating AI systems safely.
- There is a sense of urgency within these research labs to address near-term threats posed by AI.
- Offensive cyber operations, deceiving humans, and manipulating dialogue are among the harmful actions AI can already carry out.
- Collaboration between different organizations in AI research is crucial for creating approaches and standards for safe development and deployment of AI.
- Tech companies like Microsoft, Google, DeepMind, OpenAI, and ThrowBic may find ways to develop industry standards around safety and alignment despite being competitors.
- OpenAI has published thoughts on the governance of superintelligence, defining it as future AI systems more capable than AGI.
- Within the next 10 years, AI systems could exceed expert skill level in most domains and carry out significant productive activity.
- OpenAI proposes three principles for governing superintelligence: coordination among leading development efforts, international bodies regulating its use, and technical capability to ensure safety.
- Superintelligence is seen as a real possibility by many AI researchers but opinions differ on when it will occur.
- OpenAI's release of information suggests they believe superintelligence is becoming more likely to happen.
Advancements in AI and Responsible Use
34:32 - 42:07
- Samsung is able to accurately predict the level of intelligence AI models will achieve based on data and compute factors.
- There are no technical barriers to achieving super intelligence according to Samsung.
- The White House is taking steps to advance responsible AI, including an updated roadmap for federal investments in AI R&D.
- Public input is being sought on critical AI issues to mitigate risks and protect individuals.
- A report on the risks and opportunities of AI in education is being released by the US Department of Education's Office of Educational Technology.
- The White House is hosting a listening session with workers to understand how automated AI technologies are being used by employers.
- The current administration may be working on significant initiatives related to AI that have not been announced yet.
- A fake image created with the help of artificial intelligence caused disinformation and briefly impacted the stock market.
Verification Challenges and Language Preservation
41:38 - 48:55
- Synthetic content is spreading quickly and markets are reacting, but there are no clear solutions to verify content.
- Education and relying on verified sources of information are important in a world where social media is a major news source.
- Meta's massively multilingual speech project aims to promote and preserve endangered languages by using sophisticated speech recognition technology.
- Anthropic raised $450 million in series C funding to compete with OpenAI's models and build a more powerful foundational model.
- Anthropic's AI assistant model has a large context window and is highly anticipated in the AI community.
- Meta has been leading the way in open-source research and expanding language understanding.
Advancements in Robotics and Cryptocurrency
48:29 - 55:56
- Figure, a company building humanoid robots, aims to pair AI with physical robotics for advanced humanoid robots.
- Figure, a company building humanoid robots, raised $70 million in series A funding to accelerate robot development and manufacturing.
- World Coin, a cryptocurrency project by Sam Altman, raised $115 million in series C funding for online identity verification using iris scanning.
- World Coin could potentially play a role in universal basic income and the distribution of profits from AGI companies.
- Neuralink received FDA approval for human testing, bringing Elon Musk's brain-computer interface closer to reality.
- Nvidia exceeded Wall Street analysts' sales forecast and is on track to become a trillion dollar company due to their dominance in GPU chips.
NVIDIA's Earnings and Dominance
55:29 - 59:36
- NVIDIA had a historic earnings call and gained $200 billion in market cap this week.
- NVIDIA is the sixth largest company in the world, bigger than Apple, Google, Microsoft, Amazon, and Saudi state oil giant Saudi Aramco.
- The stock was up 28% in after-market trading on Wednesday after the earnings call.
- NVIDIA has been a recommended AI stock for years due to its powerful chips used in video gaming and machine learning models.
- The demand for NVIDIA's GPUs is increasing with the rise of generative AI.
- Startups are racing to raise money to buy GPUs from NVIDIA for their AI projects.
- NVIDIA is more than just a chip company; they are expanding into building models as well.