- Horizon AI
- Posts
- Sam Altman Reflects on OpenAI and the Future of AI 🔮
Sam Altman Reflects on OpenAI and the Future of AI 🔮
+ AI Poses a Danger to Financial System, Regulators Warn
Gina Acosta
December 15, 2023
Welcome to another edition of Horizon AI,
Days after being named CEO of the Year, Sam Altman gives an interview in which he discusses AGI, the brief controversy at OpenAI, and shares his predictions on the future of AI.
US regulators flagged the risks posed by AI for the first time in its annual financial stability report, cautioning about this rapidly evolving technology.
Let's jump into it!
Read Time: 3.0 min
Here's what's new today in the Horizon AI
Sam Altman Reflects on OpenAI and the Future of AI 🔮
AI Poses a Danger to Financial System, Regulators Warn
AI Research: A control method for Superintelligent AI
Trending AI Tool of the day: Caspa AI, the future of eCommerce photography
AI Image of The Day🎨: Portraits of comic characters in the style of Utagawa Kuniyoshi
The Latest in AI and Tech 💡
AI News
OPENAI
Sam Altman Reflects on OpenAI and the Future of AI 🔮
Source: TIME
2023 was a standout year for AI in the public eye, though not without its share of challenges. Just days after being named CEO of the year, Sam Altman gave an interview to TIME where he reflected on his brief departure from OpenAI and the potential risks and rewards of the technology in the future. Here are the key takeaways:
Altman's brief removal as OpenAI CEO in November proved a learning experience for him and the company, unifying the team and showing AI development must have strong governance
Altman believes artificial general intelligence (AGI) will be the most powerful technology ever created, democratizing access to information globally
However, AI risks contributing to disinformation as technology advances, especially with the 2024 US elections nearing. Altman worries AI could create highly personalized, persuasive fake content that sways public opinion
Still, he states that responsibly deployed AI has the potential to make the world "abundant and better every year"
Altman forecasts AI continuing to rapidly transform society this decade
AI REGULATION
AI Poses a Danger to Financial System, Regulators Warn
Source: Getty Images
The Financial Stability Oversight Council (FSOC) has highlighted artificial intelligence (AI) as a potential new risk in its latest annual stability report. The FSOC comprises top U.S. financial regulators, including the Federal Reserve and Treasury Department. This marks the first time the council has warned about oversight gaps and challenges posed by rapidly advancing AI tools.
Details in the report:
AI could create risks around model accuracy, bias, cybersecurity, and privacy
The opacity and complexity of some AI makes effective monitoring difficult
Recommendation for regulators and firms to deepen expertise and capacity to monitor AI innovation, usage, and potential risks
Not all aspects are negative, as the council acknowledges AI's potential to drive innovation or improve efficiencies in financial firms. It now falls upon the firms to take the necessary steps to effectively oversee complex AI tools.
AI RESEARCH
A control method for Superintelligent AI
Source: OpenAI
OpenAI researchers made an important step toward controlling superintelligent AI systems. They demonstrated a new "alignment" technique using two AI models:
GPT-2 was trained to supervise and give instructions to the much more powerful GPT-4
Despite GPT-2's weak performance and poor instructions, GPT-4 was able to correctly perform tasks on its own
This shows GPT-4 generalized from the weak supervision to find implicit knowledge needed for the tasks
Open AI believes Superintelligence—AI vastly smarter than humans—could be developed within the next ten years and this study is a step in the right direction for humans to reliably steer and control superhuman AI systems so that the most advanced AI systems in the future remain safe and beneficial to humanity.
Trending AI Tool
Caspa AI, the future of eCommerce photography
Caspa AI is a marketing suite, focused on AI product imagery. Users can create and edit unique professional photos using state of the art generative AI technology, in minutes.
Main Features:
Create and edit lifestyle photos with your products: Add photorealistic human models, animals and custom backgrounds.
Customize image templates, infographics, and more in the studio editor: add products, write text, resize images, draw inspiration from stock photos, and speed up the process with AI.
Customize AI stock photos to suit your brand preferences.
Check this AI tool here
AI Image of The Day
Portraits of comic characters in the style of Utagawa Kuniyoshi
Send your AI-generated images for a chance to get published here: [email protected]
Source: @thedarkpigment on Midjourney Discord
The latest in AI and Tech
Pope Francis calls for AI regulation
Pope Francis argues that while AI brings advantages, international governance is needed to guard against technological threats to humanity. He calls for a legally binding global AI treaty to uphold moral values as algorithms grow more advanced. The pope warns of an impending "technological dictatorship" if action isn't taken.
Grimes launches AI toy ‘Grok’
Source: Getty Images
Musician Grimes is getting into the AI toy game with a new line of plush dolls powered by conversational AI. Dubbed Gabbo, Grem, and Grok, the cute and cuddly creations can chat and bond with kids. Developed in partnership with Curio and OpenAI, Grimes herself voices the trio of smart softies. The toy Grok is a shortening of the word “Grocket,” and has no relation to her former partner Elon Musk’s AI chatbot.
GPT-4.5 leaks debunked
Supposed leaked documents of GPT 4.5 garnered significant attention yesterday after an unauthorized Reddit post surfaced, revealing changes in pricing and upgraded multimodal functionalities, including video and 3D capabilities. The leaks trended on X throughout the day. Nevertheless, Sam Altman ultimately denied the authenticity of that information.
That’s a wrap!