👀 When AI becomes conscious, here's how we will know!

🎨 AI Image of The Week

🧠 AI and Machine Learning

Germany doubles funds for AI ‘made in Europe’ (2 minute read)
The German government has revealed its plan to significantly boost funding for artificial intelligence research. The financial commitment to the advancement of AI systems now stands at nearly €1 billion, a substantial step forward. The Federal Ministry for Education and Research emphasized the pivotal role of AI as a "key technology" with boundless potential for scientific advancement, economic growth, competitiveness, and societal enrichment. Ensuring "technological sovereignty in AI" was underscored as a priority, with Germany and Europe aiming to take the reins in a world that's becoming increasingly "AI-powered." This announcement places Germany on par with the funds allocated by the EU itself, which has dedicated €1 billion annually to AI research through the Horizon Europe initiative. Europe’s AI investments still fall far behind other parts of the world. While pinpointing precise figures for China's public spending on AI remains challenging, the nation invested $13.4 billion (€12.35 billion) in private AI initiatives in 2022, a notable amount but still lagging far behind the US, which surged ahead with a whopping total of $47.4 billion (€43.4 billion).

If AI becomes conscious: here’s how researchers will know (3 minute read)
The concept of artificial intelligence gaining consciousness, a notion often seen in science fiction like HAL 9000 from the classic film "2001: A Space Odyssey," is gradually shifting from fantasy to reality, an evolution acknowledged even by AI leaders. Last year, Ilya Sutskever, the chief scientist at OpenAI, the creative force behind ChatGPT, tweeted about the potential for certain advanced AI networks to be "slightly conscious." While most experts maintain that AI hasn't yet reached the realm of consciousness, the rapid strides in AI development have prompted contemplation: how could we discern if they did? In response, a collaborative team of 19 neuroscientists, philosophers, and computer scientists has devised a checklist of criteria. This guide, unveiled on the arXiv preprint repository, lays out markers that, if satisfied, would suggest a significant likelihood of AI possessing consciousness, even though it's still awaiting peer review.

💼 Business

Maybe Bing isn't trying to compete with Google after all (4 minute read)
In the early months of 2023, Microsoft's Bing search engine was showered with accolades, thanks to its turbocharged artificial intelligence prowess. The buzz was infectious: "Bing Chat actions could be a game changer," they exclaimed. And amid these enthusiastic sentiments, someone even dared to coin the term 'Bing-aissance' without flinching. Fast forward six months, and the Wall Street Journal arrives to deflate the hype, deeming the new AI-powered Bing chatbot "cute, but not a game-changer." Lance Whitney of ZDNET chimed in, suggesting that the flashy new Bing features had failed to spark a revival in Microsoft's search engine market share. In an illustrative twist, he cited the same numbers that had appeared in the WSJ: According to market intelligence firm Statcounter, Bing's slice of the global search engine pie shrank from 3.03% in January to 2.99% in July, perking up slightly from 2.76% in April. Complementing this insight, Similarweb's data showed Bing's share as a consistent 3.23% in June, remaining virtually unchanged since the year's onset.

Meta to launch AI model for writing computer codes (2 minute read)
Meta Platforms, formerly Facebook, announced its foray into the realm of artificial intelligence by unveiling a new AI model called Code Llama, aimed at aiding in computer code composition, underscoring the company's expanding interest in cutting-edge technology. With Code Llama, freely accessible, writing code becomes a breeze as the AI responds to human text prompts, delivering code completion and debugging assistance, as elaborated in a recent blog post by the social media titan. This move aligns with the trend set by OpenAI's ChatGPT, which captivated enterprises and businesses, drawing them into the burgeoning generative AI market for enhanced capabilities and streamlined business operations. Meta's energetic AI ventures this year include the release of several free AI models, such as Llama, an open-source language model, posing a significant challenge to proprietary models from tech giants like Microsoft-backed OpenAI and Alphabet's Google.

C3.ai Stock Faces Setback as AI Hype Wanes, Shares Plummet 11% (3 minute read)
In a surprising twist, C3.ai encountered a sharp stumble in its stock value, tumbling over 11% on Thursday. This slide has sent the shares tumbling to a new low of $28.18, marking their lowest point since May 25th. Zooming out, a more extensive perspective reveals a significant tumble of over 41% from its peak earlier this year. Amid the buzz surrounding artificial intelligence (AI), there's been a hiccup in the story as C3.ai's stocks took a noteworthy nosedive. While Nvidia dazzled with impressive financials, riding high on the wave of AI chip demand, concerns about the industry's trajectory have emerged, casting a shadow over the AI arena. Interestingly, this unease has manifested itself in the slide of shares belonging to prominent players like AMD, Microsoft, AITX, and of course, C3.ai. The fortunes of C3.ai's shares have been on a downhill trajectory since their zenith in June of this year. This trend can be primarily attributed to apprehensions about the company's valuation and growth potential in the evolving AI landscape. Despite the allure of AI, C3.ai's performance has been marred by a series of annual losses, a backdrop that's hard to ignore.

❓ Misc

The New York Times prohibits AI vendors from scraping its content without permission (2 minute read)
In early August, The New York Times made changes to its terms of service (TOS) to explicitly prohibit the scraping of its articles and images for the purpose of AI training, according to a report by Adweek. This move comes at a time when the utilization of unauthorized web data for enhancing AI language applications like ChatGPT and Google Bard has been a topic of concern. The updated terms now clearly state that The New York Times' content, including articles, videos, images, and metadata, cannot be used for training any AI model without obtaining prior written permission. In Section 2.1 of the TOS, it's emphasized that the content is intended for readers' personal, non-commercial use, and this does not encompass activities such as training machine learning or artificial intelligence systems. Additionally, Section 4.1 reinforces the message by explicitly stating that the development of software programs, including AI training, requires obtaining written consent from The New York Times.

How to help high schoolers prepare for the rise of artificial intelligence (3 minute read)
Contemplating the role of artificial intelligence in healthcare might bring to mind scenes from sci-fi dystopias, but the MIT Abdul Latif Jameel Clinic for Machine Learning in Health (Jameel Clinic) has taken a more enlightening approach. In their inaugural summer program, the clinic delved into this complex topic while educating high school students. Held from July 10 to 21, the Jameel Clinic Summer Program welcomed 51 students primarily from Boston-area schools, striving for diversity and inclusion. Split into two cohorts of 25 for each week, the program featured courses like "Intro to Python," "Intro to Clinical AI," and "Intro to Drug Discovery," complemented by visits to local institutions including the Museum of Science Boston, Massachusetts General Hospital, Janssen Pharmaceuticals, and Amgen.

🐥 Best of Twitter

Thanks for reading, if you enjoyed, tell your friends!