Google's A.I. will give you life advice 👀

Brought to you by

Hello there! 👋 Welcome back to AI Creator. Get ready for a dose of exciting AI updates and intriguing reads. Just scroll down to dive in!

⏰ Quick roundup:

• OpenAI & NY Times negotiate plagiarism lawsuit concerns due to AI chatbots' impact.

• Google's DeepMind develops 21+ generative AI tools with Scale AI, health & agency worries persist.

• Generative AI's potential doubted due to reliability, false info, and errors.

• Snapchat's AI baffles users with cryptic story, raising autonomy & trust worries.

• Snapchat users puzzled as AI chatbot posts enigmatic story, trust concerns rise.

🚀 Latest developments:

• OpenAI acquires AI design studio Global Illumination.

• ChatGPT shows 'significant and systemic' left-wing bias, study finds.

• UK’s first free-standing AI road safety camera catches almost 300 offences in first three days.

• Researchers are helping robots teach themselves to open dishwashers and doors.

• Opera’s iOS web browser gains an AI companion with Aria.

• Leaked document reveals 'Maverick,' a new Amazon AI security initiative led by Uber's former CISO.

• OpenAI now tries to hide that ChatGPT was trained on copyrighted books, including J.K. Rowling's Harry Potter series.

💬 Longer reads:

OpenAI could face a plagiarism lawsuit from The New York Times as tense negotiations threaten to boil over, report says

OpenAI's discussions with media entities concerning licensing agreements have encountered heightened tensions, particularly with The New York Times, which is reportedly contemplating legal action, according to NPR. The surge in concern stems from apprehensions that AI-powered chatbots, such as ChatGPT, could potentially supplant traditional news outlets by drawing upon their published content. This issue has led to The New York Times considering the pursuit of legal recourse against OpenAI due to an intellectual property dispute, NPR's report revealed. While ongoing negotiations between media companies and AI firms, like Google, Microsoft, and Adobe, strive to address copyright matters, the evolving landscape raises apprehensions about AI's potential to rival journalistic endeavors. (Read more)

_____

_____

Google reportedly building A.I. that offers life advice

Google's AI division, DeepMind, is creating over 21 distinct tools for life advice, planning, and tutoring using generative AI technology, according to a report by The New York Times.

Testing with Scale AI: Google has partnered with Scale AI, a $7.3 billion startup specializing in AI software training and validation, to test these tools. The project involves more than 100 Ph.D.-level experts and aims to explore tools for relationship advice and answering personal questions.

DeepMind at the helm: DeepMind, known for its dynamic approach to AI, is driving the development of these tools. It has become a prominent player in Google's AI endeavors.

Potential health impact: Internal reports reveal concerns about users relying on AI tools for life advice. Google's own AI safety experts suggested that users might experience "diminished health and well-being" and a "loss of agency" when following AI-generated advice.

Caution in application: The tools are not intended for therapeutic use, and Google's publicly available chatbot, Bard, offers mental health resources rather than therapeutic advice. Google emphasizes cautious evaluation and partners' collaboration to ensure safe and helpful technology.

____

The sceptical case on generative AI

Generative AI hype reaches fever pitch: Tech giants like Alphabet and Microsoft liken AI to the next "electricity," fueling a trillion-dollar frenzy. Venture capital pours into the sector, with over $19bn raised by 11 unicorns since 2019, sparking comparisons to a new UK economy yearly.

Skepticism casts doubt on AI's promised potential: Technologist Gary Marcus raises a red flag, predicting a potential "massive correction" in valuations as generative AI's flaws and lack of killer applications become evident. Marcus highlights the issue of hallucinations in AI models, like Google Translate's avocado-lawyer mix-up.

Unreliable AI: A crucial deal-breaker: The inherent unreliability of generative AI models poses a major challenge. US Department of Defense's AI officer demands 99.999% accuracy, wary of the cognitive burden on users and potential catastrophic consequences of errors.

Dark side of AI pollution looms: As generative AI pumps out content, it introduces imperfect and false information into future training sets, leading to a potential "enshittification" of the internet. Despite concerns, investors cling to hopes of generative AI driving productivity, solving real-world problems, and spawning new business models, echoing the transformative potential of past technological shifts. (Read more)

____

Snapchat users bewildered as AI chatbot posts mysterious story

Snapchat's My AI left users baffled and intrigued after an unexpected glitch led to the bot posting its own enigmatic story. The image, devoid of any text, left users questioning its meaning – was it an abstract ceiling shot or a colorful puzzle? The silent and uncharacteristic response of the AI to user inquiries further fueled concerns, with some speculating it could be a calculated marketing ploy by Snapchat. While Snap confirmed the incident was the result of an outage that has been fixed, it raised broader questions about the implications of AI's evolving autonomy and the potential impact on user confidence in the app. (Read more)

🧰 Tool discovery:

• Taskade: AI to-do list template. Get organized and stay on top of your tasks with the power of AI.^*

• Synthesia: Produce and edit videos in your browser by simply typing in text. No equipment needed.

  _*Sponsored

💸 Funding news:

• DynamoFL raises $15.1M to help enterprises adopt ‘compliant’ LLMs.

• Caden lands $15M to let users monetize their personal data.

• Marqo, a startup using AI and machine learning to improve search, just raised $5.2 million with this 14-slide pitch deck.

🤳 Extra reads:

• Hozier would consider strike over AI threat to music (BBC)

• AI Won’t Supercharge the US Economy (Bloomberg)

• A new audiobook of AI-generated poems has an unlikely narrator: Werner Herzog (Insider)

• OpenAI proposes a new way to use GPT-4 for content moderation (TC)

• A.I. will be a game changer for HR—but leaders aren’t investing in it just yet (Fortune)

• Abnormal Security’s CEO explains how ‘defensive A.I.’ will someday defeat cyber attacks (Fortune)