Data and AI newsletter: June 2023
We are thrilled to present the latest edition of our data and AI newsletter, packed with exciting updates and industry insights. Although it arrives a little later than usual, it's well worth the wait! The past few weeks have been an exhilarating whirlwind of events, including our participation in London Tech Week 2023, the Credit Risk Management Conference, and our webinar on AI Trends in Banking and Fintech. If you missed the webinar, don't worry! Simply drop us a line at [email protected], and we'll be delighted to share the recording with you.
Now, grab a cup of coffee, settle in, and join us as we delve into the world of data and artificial intelligence!
NEWS AND INSIGHTS
Artificial intelligence has pros and cons, but one thing is for sure - AI is here to stay. EU Parliament is ready to negotiate the AI Act, the world’s first AI law, to ensure that Europe can continue to innovate while protecting people. Under the proposed legislation AI technologies would be classified into 4 different categories of risk, from minimal to unacceptable.
McKinsey & Company published a great write-up on the economic possibilities of GenerativeAI.
Their research estimates that generative AI could add the equivalent of $2.6 trillion to $4.4 trillion annually… Great! This is a big opportunity, but how does it apply to you? Read the full article here to see all of McKinsey & Company’s findings.
At Microsoft's Build 2023 conference, several exciting announcements were made.
Windows Copilot was introduced, making Windows 11 the first PC platform to provide centralized AI assistance. This AI-powered feature aims to help users easily accomplish tasks and increase productivity.
A new generative AI tool for Adobe Photoshop will enable users to swiftly extend photos and add or delete objects using text prompts. Generative Fill is now available in beta, but Adobe claims that a full version of it for Photoshop will come later this year.
Sam Altman says the research strategy that birthed ChatGPT is played out and future strides in artificial intelligence will require new ideas.
”I think we're at the end of the era where it's going to be these, like, giant, giant models,” he told an audience at an event held at MIT earlier. Read the full story at wired.com
Swiss Researchers used artificial intelligence to help Gert-Jan Oskam, who had been paralyzed from the waist down for more than ten years, regain control over his lower body.
Following a motorcycle accident in 2011, Gert-Jan Oskam has been paralyzed from the waist down. However, there has been a recent breakthrough that has enabled him to regain mobility. Scientists have successfully developed a "digital bridge" that connects Oskam's brain and spinal cord, effectively bypassing the damaged areas.
Through the use of an AI thought decoder, the researchers were able to capture Oskam's thoughts and translate them into spinal cord stimulation, thereby restoring voluntary movement.
INTERESTING AI STARTUPS AROUND
Sensoneo, a Bratislava-based smart waste management startup, secures €6.2M Series A funding to expand its solution to other regions, including APAC, and strengthen its global position. Read more at therecusive.com.
It uses Google’s multimodal foundation models as APIs including PaLM, Imagen, Codey, and Chirp. Runs on Vertex AI.A simple, easy-to-use interface for prompt design, tuning, and deployment.
💡 Generative AI learning path now available
“This learning path guides you through a curated collection of content on generative AI products and technologies, from the fundamentals of Large Language Models to how to create and deploy generative AI solutions on Google Cloud.”
💡 More-efficient approximate nearest-neighbour search - Amazon Science
In a paper by the Amazon Science team, presented at this year’s Web Conference, they described a new technique that makes the graph-based nearest-neighbour search much more efficient. The technique is based on the observation that, when calculating the distance between the query and points that are farther away than any of the candidates currently on the list, an approximate distance measure will usually suffice. Accordingly, they propose a method for computing approximate distance very efficiently and show that it reduces the time required to perform an approximate nearest-neighbour search by 20% to 60%.
Earlier this year, we had the pleasure of hosting a captivating PyData Sofia Meetup, featuring our esteemed colleague, Georgi Nalbantov, PhD. In his insightful talk on "The Bias-Variance Trade-Off," Georgi shared valuable insights and practical examples. He has recorded the entire talk and made it available on his YouTube channel. ▶️ Watch the video here.