AI and Societal Responsibility: Should GPT-4 Be Let off the Leash?

OpenAI aims to limit the potential societal destruction and danger in the wrong hands.


Sam Altman Asks: Should OpenAI let GPT-4 Off the Leash?

The article discusses the responsibility of OpenAI, an organization that aims to limit the potential societal destruction caused by artificial intelligence (AI) while advancing the technology. OpenAI has developed ChatGPT, a language model that imitates human consciousness, but also recognizes the potential danger of AI technology in the wrong hands. After eight months of regulation, they released the GPT-4 model to the public. OpenAI’s responsibility in limiting societal destruction caused by AI is significant, but governments and the private sector must also adapt to this technology quickly. OpenAI co-founder Sam Altman believes that acknowledging the possibility of super-powerful AI machines killing humans is important to solve the problem, and there needs to be a tight feedback loop and democratic decision-making around this technology. While OpenAI has faced criticism for not being open enough with their technology, they have widely distributed it.


Elon Musk Signs Open Letter Urging AI Labs to Pump the Brakes

Elon Musk and over a thousand tech experts have signed an open letter urging leading AI labs to pause the training of new systems for six months. Recent advances in AI have presented “profound risks to society and humanity,” the letter says, adding that “Advanced AI could represent a profound change in the history of life on Earth and should be planned for and managed with commensurate care and resources.” The group is calling for a six-month pause for AI safety research, which should be implemented with safety protocols rigorously audited and overseen by independent outside experts. The letter highlights the urgent need to accelerate alternative AGI architectures that are safe and formally verifiable by design.

Here is a video I made that goes over it in detail:


Baidu’s AI Chatbot Ducks Out of Public Demo

Baidu’s new AI chatbot, ERNIE, has disappeared from public view, after the company canceled a Monday product launch where it promised to unveil the bot’s cloud services and application products. The event was transformed into a closed-door meeting for the first batch of invited testing enterprises. Baidu faced criticism after admitting that the demos shown were recorded and not live, as the bot’s mental gears were uninteresting to watch. The Chinese firm’s share price dipped on Monday, but has since rebounded on news that ERNIE will be added to Baidu’s forthcoming electric vehicles.


Covariant’s CEO on Building AI that helps Robot Learn

Covariant, a Bay Area firm, has developed an AI-based system to train network robots on better picking up objects in warehouses. According to the co-founder and CEO, Peter Chen, Covariant’s system differs from other robotics companies by going deeper into AI than anyone else. The system is not based on off-the-shelf or open-source models but is rather built on foundational models that can be applied to a vast range of circumstances. The system has a clear need, given the vast range of sizes, shapes, textures and colors of goods found in a warehouse. The company’s systems are fed into a central database where every pick informs machines on how to pick in the future. While Covariant’s focus is on picking and placing, the company’s technology has potential applications in other areas, given the convergence of AI, software and mechatronics.


Goldman Estimates that AI Could Drive $7 Trillion in Global Growth in 10 Years. Here’s How To Play It.

Goldman Sachs predicts that generative AI will drive nearly $7 trillion in global economic growth over the next decade, with productivity increasing by 1.5%. This innovation can produce new content such as text, video, images or computer code, streamlining business workflows and automating routine tasks. The total addressable market for generative AI is $150 billion, and there will be opportunities for software companies to upsell and cross-sell products with AI. Goldman recommends investing in tech giants such as Microsoft, Salesforce, and Adobe, who are leaders in the field and are incorporating generative AI into their products.


OpenAI CEO Sam Altman Reveals One Job that ChatGPT can Replace Soon

In a recent podcast interview, OpenAI CEO Sam Altman specified that customer service is a category of job that can be “massively impacted” by ChatGPT. This statement adds to concerns about the AI chatbot’s potential to replace human jobs. However, Altman also acknowledged that human beings have limitless creativity and will eventually find new things to do. ChatGPT has already been utilized by various companies for various purposes, and schools are also starting to warm up to the idea of using the chatbot for educational purposes.


Altman vs Musk: OpenAI Treads on Tesla’s Robot Turf with Investment in Norway’s 1x

OpenAI and Tesla co-founders, Sam Altman and Elon Musk, are once again at odds as OpenAI invests in Norwegian robotics company 1X, which builds wheeled, humanoid robots named EVE and is preparing to unveil a bipedal android named NEO. The funding will support NEO’s development and “scale manufacturing” of EVE in Norway and North America, where security firm ADT already plans to deploy it for night patrols. This move may put OpenAI in indirect competition with Musk’s Tesla, which is promising to launch a bipedal humanoid robot called the Tesla Bot or Optimus by mid-2021.


Cerebras Systems Releases Seven New GPT Models Trained on CS-2 Wafer-Scale Systems

Cerebras Systems, a pioneer in artificial intelligence compute for generative AI, has trained and released a series of seven large language models (LLMs) for open use by the research community. This is the first time a company has used non-GPU based AI systems to train LLMs up to 13 billion parameters and is sharing the models, weights, and training recipe via the industry standard Apache 2.0 license. All seven models were trained on the 16 CS-2 systems in the Cerebras Andromeda AI supercomputer. By releasing the models under an open-source license, Cerebras is promoting open research and access, as well as enabling anyone to create powerful, industry-specific applications with minimal work. The models released today make a number of technical contributions, including the demonstration of a simple, data-parallel only approach to training, which requires no code or model modification to scale to arbitrarily large models.