Not A Bot
Posts
💰️ AI could make countries rich

💰️ AI could make countries rich

PLUS: ChatGPT aids social equity in India's poorest regions

Haroon Choudery
June 16, 2023

New here? Welcome to NAB, and make sure to subscribe for more!

Happy Friday, fellow humans. 👋

CAVA had a wildly successful IPO. Beyoncé is back to taking over the world. Fed paused interest rate hikes as US inflation comes down. AI is poppin’ off more than ever.

Are we officially back?

Let’s dive into it…

Have anything you’d like to share with over 40k subscribers?

Sponsor Not a Bot

🧵 In today's edition:

💿️ Graphic chip wars continue
🤖🧑 Yann LeCun’s vision for human-like AI
💰️ AI could make countries rich
🤑 AI Fundraising News

🤖 Top AI News

💿️ Graphic chip wars continue

Graphic chips are the ball and chain that keeps the generative AI space from absolutely exploding.

Companies need tens of thousands of advanced graphics chips to train large AI models that can generate original text and analysis, but their demand far outweighs their supply.

Nvidia is currently the king at chip production, with AMD, Intel, and other chipmakers trying to catch up - but the gap between the companies is only getting wider.

Nvidia’s biggest rival, AMD, unveiled a new AI chip, the MI300X, but its release is scheduled for the first half of 2024, which gives Nvidia a considerable 18-month lead in the market.

AMD claims that the chip outperforms Nvidia's flagship H100 but could not point to potential customers. Plus, its new chip will only handle AI inferencing and not the more demanding task of training.

What does this mean? For now, the booming generative AI chip market still belongs to Nvidia.

Read more: Financial Times

🤖🧑 Yann LeCun’s vision for human-like AI

Yann LeCun, one of the three "godfathers of AI" and Meta’s chief AI scientist, has been quite active recently.

This week, he released a first look into the “new architecture” his team has developed that aims to overcome “key limitations of even the most advanced AI systems today.”

His goal is simple: to “create machines that can learn internal models of how the world works so that they can learn much more quickly, plan how to accomplish complex tasks, and readily adapt to unfamiliar situations.”

You can read the paper here or catch it presented at CPVR 2023 next week.

LeCun also provided some thoughts on the AI x jobs debate, stating that while computers will eventually surpass human intelligence, they won't threaten humanity or the job market.

He dismissed fears of harmful AI and compared it to people's anxiety over turbo jets in the 1930s, which were eventually made "incredibly reliable and safe."

An interesting comparison, considering my brain still goes to the worst-case scenario whenever there’s a little turbulence 😬 🛫

LeCun added that regulation could be necessary, but each AI application would require its own set of rules.

Honestly, we can speculate all we want, but time will tell all.

Read more: BBC

💰️ AI could make countries rich

McKinsey released a 68-page (read: long) report on the future of generative AI and some thoughts on how it could affect the workforce’s next “wave of productivity.”

Here are some highlights:

AI is poised to add 2.6 trillion to $4.4 trillion to the global economy. For comparison, in 2021, UK’s GDP was $3.1 trillion, India’s was $3.2 trillion, and Australia’s was $1.6 trillion. Meaning - AI can make countries a LOT of money.
Out of 63 AI use cases examined, 75% of generative AI’s value creation falls under Customer operations, marketing and sales, software engineering, and R&D.
Current generative AI tech has the potential to automate activities that “absorb 60 to 70 percent of employees’ time today.”

The report goes further into how AI can help change the banking, pharmaceutical, and medical-product industries, as well as its widespread societal impacts.

Give it a read if you have some time, or simply google “AI summary tools” and feed it the report for a somewhat quicker read :)

Read more: McKinsey

🤑 AI Fundraising News

Blackbird.AI Raises $20M in Series B Funding to provide a constellation platform for organizations to detect and respond to threats that cause reputational and financial harm in real-time.

Deepchecks Raises $14M in Seed Funding to provide an open-source platform that enables practitioners, developers, and other stakeholders to holistically test and monitor machine learning models and datasets and gain visibility and confidence in every step of the process.

🗞️ Byte size: AI article summaries

Disclaimer: AI is (partially) used to summarize these articles.

Google, maker of AI chatbot Bard, warns its employees about using chatbots [Mashable] - Google has warned its employees not to share confidential information with any AI chatbot, including its creation called Bard. The company is concerned about the potential privacy issues and data sharing associated with AI language models. Google has updated its Bard privacy notice page with a warning not to include identifiable information in the conversations with Bard. The warning also applies to users of other AI chatbots, who may not be aware that human workers can access and review their chats' content. (Read more)

A Chatbot That Won't Take Bribes for Giving Advice Is a Hit in India [Yahoo Finance] - ChatGPT is being used in India to help the country's poorest people access government anti-poverty programs without getting snarled in red tape and corruption. Trials in Bangalore and Mewat in northern India show that AI can aid social equality and assist in professional communications. The AI tests are also being built to help the underprivileged seek legal justice, dispense advice to farmers, and help migrant workers get support in cities. However, while AI can bridge chasms like language and literacy, it can also exacerbate the divide by excluding those without access to technology. (Read more)