April 12, 2023
AI News: Rundown of the key events of the past weeks
Generative AI could raise global GDP by 7%, Goldman Sachs report says
According to researchers at Goldman Sachs, in the next ten years, generative AI can drive a 7% increase in global GDP and significantly boost productivity.
The researchers estimate that generative AI could impact up to 25% of current employment, or some 300 million full-time jobs globally. However, jobs displaced by automation have historically been offset by creating new roles. These new occupations then account for most of the long-term employment growth.
As Goldman Sachs researchers point out, this process of job displacement due to technological leaps has been historically repeating. For example, 60% of today’s workers are employed in occupations that didn’t exist in 1940.
So although we are undoubtedly about to face a shift in how we work, this process has been at the heart of our growth and development for centuries.
The debate about generative AI is heating up.
An open letter signed by prominent tech leaders, including Elon Musk and Steve Wozniak, has stirred up a debate about the current AI developments. The letter calls for a halt in training AI systems that exceed the capabilities of OpenAI's recently launched GPT-4 for at least six months. This pause should allow evaluation of these advanced technologies' ethical, societal, and security implications.
The responses to the letter vary, with many supporting the need for further assessment of the impact of generative AI. However, questions also arise about how such a pause could be enforced and what benefits it could bring to those who disobey it.
Compliance of ChatGPT with data privacy is becoming a MUST
The calls for data privacy laws and regulations are also getting louder. Techcrunch reports that Italy has temporarily banned ChatGPT until Open AI complies with the EU's General Data Protection Regulation (GDPR). The Italian data protection agency cited several reasons for the block, including no legal basis, lack of accuracy, and no age verification or filters for users under 13.
As more countries may follow Italy’s stance, OpenAI must ensure the language model complies with all the privacy and data protection laws.
From ChatGPT to products enhanced with generative AI
AI has been a black box powering many products and services we interact with for years. However, ChatGPT has marked one of the first B2C cases where AI is the product’s core.
While it may help with personal productivity, the most economic potential lies in customizing the generative AI model for other products or software tailored for specific businesses.
(If you want to learn more about how tailored generative AI could help your business thrive, contact us at contact@blindspot.ai.)
Some global leaders are already enhancing their products with generative AI. For example, Bloomberg, a global business and financial information provider, has just launched BloombergGPT.
This “ChatGPT for Finance” is a 50 billion parameter model, trained on financial data and focused on tasks such as sentiment analysis, entity recognition, news classification, and analyzing vast amounts of data available on the Bloomberg Terminal.
This is a great use case of generative AI in business, and we expect many similar applications to pop up soon.
Another global pioneer, Adobe, unveiled Firefly, a tool that allows users to make an infinite range of custom image creations.
Others aim to challenge Google’s search engine dominance, such as Perplexity AI, a startup that has recently secured $26 million in funding. It develops a conversational answer engine drawing from credible sources in real-time.
What about the tech giants?
Google has officially launched Bard, its AI chatbot, to compete against OpenAI's ChatGPT and Microsoft's Bing. The company started with a limited public rollout, which means only U.S. and U.K.-based users can currently join a waiting list at https://bard.google.com/.
However, Bard's debut in February suffered a setback after it made a factual error during a live demo, which resulted in Google losing $120 billion in market value. Despite this setback, the launch of Bard shows Google's commitment to advancing the capabilities of AI chatbots.
Microsoft is joining the race and recently announced that its Co-pilot for Office will take virtual assistants to a whole new level. Powered by OpenAI's GPT-4, the Co-pilot should be available across all Microsoft 365 apps, including Teams, Outlook, Excel, Word, and PowerPoint.
These features are currently available in the United States by signup, but Microsoft emphasizes that they are still experimental, and accuracy and relevance cannot be guaranteed.
Looking to dive deeper? Here is a couple of recently released podcast you shouldn’t miss
🎧 Lex Fridman interviewed Sam Altman (CEO of Open AI)
Where is ChatGPT now? How is it going to evolve? And what about all those AI safety considerations? Lex Fridman, a well-known research scientist at MIT and host of a popular podcast, asked all this and more Sam Altman, CEO of Open AI.
Listen here: Sam Altman: OpenAI CEO on GPT-4, ChatGPT, and the Future of AI | Lex Fridman Podcast
Some of the key takeaways are:
- The release of GPT-4 involved significant efforts toward AI safety considerations
- The challenge of regulating speech and evaluating GPT models highlights the need for a democratic process to establish clear boundaries and rules for AI technology.
- While GPT-like models may automate programming to a large extent, great programmers will still have a fundamental human element.
🎧 Michal Pěchouček, our Member of the Board, talked about AI in a CzechCrunch podcast
In another podcast from CzechCrunch, our co-founder, Michal Pěchouček, talked about the recent development of AI and its global impact.
During the podcast, he outlines that he expects artificial intelligence to start advancing scientific research fundamentally. "Science will be automated; scientists will generate their research and results much faster." Michal also added that AI is poised to radically transform various administrative roles, such as call center jobs.
However, he believes there is no need to be concerned about so-called general artificial intelligence now.
"It may happen, there is no strong argument why superintelligence could not occur, but the path to it will still be long and complex."
Listen to the full podcast here (in Czech): https://cc.cz/umela-inteligence-zasadne-zrychli-vedu-pomaha-ale-i-zlocincum-rikaji-pechoucek-s-rehakem/
That's a wrap of our first rundown of the latest developments in AI. Make sure to subscribe to our blog to get notified about the next edition because the age of AI is here.
If you have any feedback or suggestions on what topics we should explore deeper, please don't hesitate to comment or contact us directly.
Are you interested in implementing AI in your business? Then reach out to us at contact@blindspot.ai, and let’s chat!
Until next time!
Team of Blindspot AI
