March 31, 2023

Software

Is AI Coming for Coders First?

New York Magazine, 03/31/23. The impact of AI is in many cases speculative. There may be widespread job loses, wholesale changes in job responsibilities in addition to productivity gains. However, it is a certainty that AI automation will hit software development hard.

AI assistants like GitHub Copilot are already in use to speeding up programming. Google is developing its own general-purpose coding assistant. 

While the impact of AI on labor remains uncertain, OpenAI suggests that around 80% of the U.S. workforce could have at least 10% of their work tasks affected by the introduction of LLMs. Approximately 19% of workers may see at least 50% of their tasks impacted. 

Ultimately, the effects of AI automation on labor will manifest early, if not first, in the industry where it’s first and most thoroughly deployed, and it seems especially capable in software development. READ MORE

Limitations

ChatGPT’s intelligence is zero, but it’s a revolution in usefulness, says AI expert

ZDNET, 03/31/23. Generative AI systems like ChatGPT, DALL-E, and Midjourney are poised to revolutionize many industries. While these systems are good at generating text and images, they lack the capacity to tackle challenges that are new to them. It is argued that true intelligence is the ability to solve completely novel problems which these AI models cannot do. 

Despite this limitation, these generative models will deliver significant advances to industries such as illustration and filmmaking. Here, they can produce creative and visual work efficiently and cost-effectively. 

The rise of AI models like ChatGPT and GPT-4 is a “usefulness revolution” that will accelerate technological developments. However, they are not a direct path to achieving artificial general intelligence (AGI) with human-level intelligence. READ MORE

Chatgpt

OpenAI connects ChatGPT to the internet

TechCrunch, 03/31/23. ChatGPT can now access the internet through plugins that grant it access to third-party knowledge sources and databases. 

The most intriguing plugin is OpenAI’s web-browsing plugin which allows ChatGPT to retrieve content from the web using the Bing search API and cite its sources in responses. However, chatbots with web access can be risky, as shown by past experiments like OpenAI’s WebGPT and Meta’s BlenderBot 3.0. 

To prevent undesirable behaviors, OpenAI has implemented safeguards informed by internal and external red teams. Additionally, OpenAI has released a code interpreter for ChatGPT and open-sourced a retrieval plugin that enables ChatGPT to access snippets of documents from data sources. 

Early collaborators have also built plugins for ChatGPT, including Expedia, Instacart, OpenTable, Shopify, Slack, Wolfram, and Zapier. READ MORE

Perspective

A.I. May Change Everything, but Probably Not Too Quickly

The New York Times, 03/31/23. Artificial intelligence (A.I.) is rapidly gaining influence on the economy, with its effects expected to grow significantly in the coming years. However, the economic impacts of A.I. are difficult to predict, as they depend on various factors such as the rate of technological advancement, government policies, and workers’ adaptability. 

Previous technological advances, such as the rise in computing power, took a long time to have a significant economic payoff. Similarly, the economic benefits of A.I. may take longer to materialize than expected. Nonetheless, A.I. will likely make the economy more productive, while also potentially harming workers whose skills have been devalued. 

Long-term economic projections are always subject to uncertainty. However, they matter because they underlie the long-term budget outlook and inform current policy decisions. READ MORE

regulation

This gung-ho government says we have nothing to fear from AI. Are you scared yet?

The Guardian, 03/31/23. The rise of social media has had significant impacts including the fueling hate speech to, the polarization of communities and the destabilization of democratic processes. Governments’ failure to regulate tech giants blinded by their wealth generation has left us with an uncontrollable mess. Now, with artificial intelligence (AI) becoming a reality, the world is at a crossroads again. 

On one hand, AI can bring about magical breakthroughs in medicine and productivity. On the other, it can disrupt the social and economic order at breakneck speed. Leading AI experts have called for a six-month pause to understand its implications. The British government, desperate for economic growth, argues for light regulation to get ahead of the AI race. 

As AI continues to develop, it is crucial to learn from the mistakes made with social media and consider new rights for humans living alongside AI to avoid a potentially disastrous future. READ MORE