From chatbots to superintelligence: Charting AI’s ambitious journey

Don’t miss OpenAI leaders Chevron, Nvidia, Kaiser Permanente and Capital One only at VentureBeat Transform 2024. Gain essential insights about GenAI and expand your network at this exclusive three-day event. Learn more

Is humanity on the verge of creating its intellectual superior? Some think we are on the verge of such a development. Last week, Ilya Sutskever unveiled his new startup, Safe Superintelligence, Inc. (SSI), which is dedicated to building advanced models of artificial superintelligence (ASI) – a hypothetical AI far beyond human capabilities. In a statement about the SSI’s launch, he said “superintelligence is within reach” and added: “We approach security and capability together.”

Sutskever has the credentials to aspire to such an advanced model. He was a founding member of OpenAI and previously served as the company’s chief scientist. Before that, he worked with Geoffrey Hinton and Alex Krizhevsky at the University of Toronto to develop “AlexNet,” an image classification model that transformed deep learning in 2012. More than any other, this development launched the growth of AI during last decade, in part by demonstrating the value of parallel processing of instructions by graphics processing units (GPUs) to speed up deep learning algorithm performance.

Sutskever is not alone in his belief in superintelligence. SoftBank CEO Masayoshi Son said late last week that AI “10,000 times smarter than humans will be here in 10 years.” He added that achieving ASI is now his life mission.

AGI within 5 years?

Superintelligence goes beyond artificial general intelligence (AGI), also still a hypothetical AI technology. AGI would surpass human capabilities in most tasks of economic value. Hinton believes we could see AGI within five years. Ray Kurzweil, principal AI researcher and visionary at Google, defines AGI as “AI that can perform any cognitive task that an educated human can.” He believes this will happen in 2029. Although in reality, there is no generally accepted definition of AGI, which makes it impossible to accurately predict its arrival. How would we know?

Countdown to VB Transform 2024

Join enterprise leaders in San Francisco July 9-11 for our premier AI event. Connect with peers, explore the opportunities and challenges of generative AI, and learn how to integrate AI applications into your industry. Register Now

The same can be said for superintelligence. However, at least one forecaster is on record as saying that superintelligence could arrive soon after AGI, perhaps by 2030.

Despite these expert opinions, it remains an open question whether AGI or superintelligence will be achieved within five years – or ever. Some, like AI researcher Gary Marcus, believe that the current focus on deep learning and language models will never achieve AGI (let alone superintelligence), seeing these as fundamentally flawed and weak technologies that could advance only through the brute force of more data and computers. power.

Pedro Domingos, professor of computer science at the University of Washington and author of The Master Algorithmsees superintelligence as one pipe dream. “Ilya Sutskever’s new company is guaranteed to succeed, because the superintelligence that is never achieved is guaranteed to be safe,” he posted on X (formerly Twitter).

What comes next

One of these views may be correct. No one knows for sure if AGI or superintelligence will come or when. As this debate continues, it is crucial to recognize the gap between these concepts and our current AI capabilities.

Rather than speculating only on the far-future possibilities that are fueling the stock market’s abundant dreams and public anxiety, it is at least as important to consider the more immediate developments that are likely to shape the landscape of AI in the coming years. These developments, while less sensational than AI’s grandest dreams, will have significant real-world impacts and pave the way for further progress.

As we look ahead, the coming years will likely see AI language, audio, image and video models—all forms of deep learning—continue to develop and multiply. While these advances may not achieve AGI or superintelligence, they will undoubtedly increase the capabilities, utility, reliability and application of AI.

That said, these models still face some significant challenges. A major drawback is their tendency to hallucinate or confuse from time to time, essentially creating responses. This lack of reliability remains a clear barrier to widespread adoption at present. One approach to improve AI accuracy is feedback augmented generation (RAG), which integrates current information from external sources to provide more accurate responses. Another might be “semantic entropy,” which uses a large linguistic model to check the work of another.

There is no universal answer to AI (yet)

As robots become more reliable over the next year or two, they will be increasingly incorporated into business applications and workflows. To date, many of these efforts have fallen short of expectations. This result is not surprising, as the inclusion of AI constitutes a paradigm shift. My view is that it’s still early days, and that people are still gathering information and learning how to best deploy AI.

Wharton professor Ethan Mollick echoes this view in his own A useful thing newsletter: “Right now, no one—from consultants to typical software vendors—has a universal answer for how to use AI to unlock new opportunities in any particular industry.”

Mollick argues that much of the progress in implementing generative AI will come from workers and managers experimenting with applying tools in their areas of domain expertise to learn what works and adds value. As AI tools become more capable, more people will be able to advance their work output, creating a flywheel of AI-powered innovation within businesses.

Recent advances demonstrate this innovative potential. For example, Nvidia’s Inference Microservices can accelerate the deployment of AI applications, and Anthropic’s new chatbot Claude Sonnet 3.5 is said to outperform all competitors. Artificial intelligence technologies are finding increasing application in various fields, from classrooms to car dealerships and even in the discovery of new materials.

Progress is likely to accelerate steadily

A clear sign of this acceleration came from Apple with the recent launch of Apple Intelligence. As a company, Apple has a history of waiting to enter a market until there is sufficient technological maturity and demand. This news suggests that AI has reached that tipping point.

Apple Intelligence goes beyond other AI announcements by promising deep integration across apps while preserving context for users, creating a deeply personalized experience. Over time, Apple will enable users to implicitly group multiple commands together into a single request. These can be run across multiple applications, but will appear as a single result. Another word for this is “agents”.

During the Apple Intelligence launch event, SVP of software engineering Craig Federighi outlined a scenario to show how these will work. As reported by Technology Review, “an email comes in postponing a work meeting, but his daughter is appearing in a show that night. His phone can now find PDFs of performance information, predict local traffic, and let him know if he’ll make it on time.

This vision of AI agents performing complex multi-step tasks is not unique to Apple. In fact, it represents a broader shift in the AI ​​industry toward what some call the “age of agency.”

AI is becoming a real personal assistant

In recent months there has been growing industry discussion about moving beyond chatbots and into the realm of “autonomous agents” that can perform multiple connected tasks based on a single request. More than just answering questions and exchanging information, this new set of systems uses LLM to perform multi-step actions, from software development to flight reservations. According to reports, Microsoft, OpenAI and Google DeepMind are all preparing AI agents designed to automate more difficult multi-step tasks.

OpenAI CEO Sam Altman described the vision of the agent as a “super-competent colleague that knows absolutely everything about my entire life, every email, every conversation I’ve ever had, but doesn’t feel like an extension.” In other words, a true personal assistant.

Agents will also serve applications across enterprise uses. McKinsey senior partner Lari Hämäläinen describes this advance as “software entities that can orchestrate complex workflows, coordinate activities among multiple agents, apply logic, and evaluate responses. These agents can help automate processes in organization or add workers and customers as they perform processes.

Enterprise agent-focused startups are popping up, too — like Emergence, which rightfully just came out of stealth mode. According to TechCrunch, the company claims to be building an agent-based system that can perform many of the tasks typically handled by knowledge workers.

The way forward

With the pending arrival of AI agents, we will join the ever-connected world even more effectively, both for personal use and for work. In this way, we will increasingly dialogue and interact with digital intelligence everywhere.

The path to AGI and superintelligence remains shrouded in uncertainty, with experts divided on its feasibility and timeline. However, the rapid evolution of AI technologies is undeniable, promising transformative advances. As businesses and individuals navigate this rapidly changing landscape, the potential for AI-driven innovation and improvement remains huge. The journey ahead is as exciting as it is unpredictable, with the lines between human and artificial intelligence continuing to blur.

By taking proactive steps now to invest in and engage with AI, improve our workforce, and consider ethical considerations, businesses and individuals can position themselves to thrive in the AI-driven future.

Gary Grossman is EVP of the technology practice at Edelman and global leader of the Edelman AI Center of Excellence.

Data DecisionMakers

Welcome to the VentureBeat community!

DataDecisionMakers is where experts, including technical people who do data work, can share insights and innovations about data.

If you want to read about the latest ideas and up-to-date information, best practices and the future of data and data technology, join us at DataDecisionMakers.

You might even consider contributing an article of your own!

Read more from DataDecisionMakers

#chatbots #superintelligence #Charting #AIs #ambitious #journey
Image Source :

Leave a Comment

Your email address will not be published. Required fields are marked *

Scroll to Top