Last week, I went on an exciting adventure thanks to Google’s Bard chatbot. I rode the Bustang bus from Colorado Springs to Denver, and in the process, I discovered Bard was horribly, majestically, and terribly wrong about public transportation. What a mess!
And! Plus! I saw a human being tackled and otherwise pummeled by two other humans at a Bustang stop in downtown Colorado Springs. Exciting! More on that in another, future Substack post.
And so that brings up this week’s Keepin’ Up with the Chatbots. There’s lots of news and updates this week. Google is front and center of the latest AI and chatbot news as the data giant continues gobbling up all things related to information. Bots ain’t what you used to think and they have evolved to “scraping intelligence” from chatbots. Plus college students and ChatGPT continue learning from one another. And more!
Gemini’s Multmodal Capabilities: A Game Changer in AI
👉 Google's newest AI model, Gemini, is a true powerhouse. This versatile, multimodal system is designed to handle diverse data types with ease, making it a potential game-changer in the world of AI. Unlike its predecessors, Gemini doesn't rely on specialized models, and it effortlessly adapts to various tasks. But that's just the tip of the iceberg.
Drawing inspiration from AlphaGo and harnessing language capabilities akin to ChatGPT, Gemini is a fusion of cutting-edge techniques. Early demonstrations have left experts astounded, hinting that it could outperform OpenAI's GPT-4 by a staggering factor of five. This exceptional performance is attributed to Gemini's access to Google's extensive reservoir of consumer product data.
As Gemini prepares for its grand entrance, the AI landscape braces for a seismic shift. The release of this remarkable AI model is poised to ignite fierce competition and catalyze groundbreaking advancements in the field of artificial intelligence.
👉 AI models are surging in popularity, yet they face a growing threat from bots employing novel and questionable tactics to "extract intelligence." Guillermo Rauch, CEO of Vercel, recently uncovered this new breed of bot that scavenges data from AI models like OpenAI's GPT-4.
The surge in bot activity stems from the demand for high-quality data to fuel AI models and the escalating costs associated with employing top-tier models. Also, some exploit bots to circumvent China's AI model censorship. While Vercel provides protective technology against these bots, SaaS enterprises reliant on AI models may encounter financial jeopardy in the face of these unscrupulous agents.
Amazon to Invest Up To $4 Billion in AI Startup Anthropic
👉 In a significant development, Amazon has unveiled its ambitious intention to inject a substantial $4 billion into the burgeoning AI startup, Anthropic. This strategic investment is set to propel Anthropic towards the creation of a groundbreaking "frontier model" named "Claude-Next," a marvel projected to outshine the potency of even the most formidable AI currently in existence by a factor of ten.
Leveraging Amazon's robust cloud computing services and harnessing the power of AWS Trainium and Inferentia chips, Anthropic aims to mold and deploy its cutting-edge models with unprecedented proficiency. In a mutually beneficial aspect of the deal, Amazon secures an exclusive gateway to Anthropic's AI-driven text analysis chatbot, Claude, further solidifying its position in the ever-evolving AI landscape.
You Can Now Prompt ChatGPT With Pictures and Voice Commands
👉 OpenAI's ChatGPT chatbot has introduced new features that allow users to use voice commands and images for prompts. The voice chat feature functions similarly to virtual assistants like Alexa or Google Assistant, where users can speak their questions and receive spoken answers. OpenAI has developed a text-to-speech model that generates human-like audio from text and a few seconds of sample speech. Users can choose from five voice options.
OpenAI acknowledges the risks associated with synthetic voices, such as impersonation and fraud, and plans to control and restrict its use. The image search feature allows users to upload photos and ask questions related to the image, utilizing ChatGPT's back-and-forth interaction for refining answers. OpenAI has limited the bot's ability to analyze and make direct statements about people for accuracy and privacy reasons.
Nearly A Third of College Students Used ChatGPT Last Year According to A Survey
👉 Recent survey findings have unveiled a notable trend among college students, revealing that nearly one-third of them embraced the use of ChatGPT in the past year. Interestingly, the surge in adoption comes in spite of a significant portion of students who harbor reservations, viewing its utilization as a potential conduit for academic dishonesty akin to cheating or plagiarism. ChatGPT's allure among the student demographic lies in its versatile capabilities, ranging from essay generation to math problem-solving and on-the-fly information summarization.
But this burgeoning reliance on ChatGPT has ignited a complex debate within academic circles. Educational institutions in Ohio and beyond grapple with the challenge of formulating appropriate responses to college students’ use of ChatGPT and chatbots in general. Some colleges and institutions have introduced stringent guidelines that permit ChatGPT usage solely under explicit instructor consent, whereas others have taken a more conservative stance by banning it altogether, driven by mounting concerns regarding its capacity to facilitate academic impropriety.
👉 Everyday, there’s more email notifications sent to me of breaking news related to AI and the chatbots. It’s hard keeping up with all the changes, and that’s why I put together this Keepin’ Up with the Chatbots post — so you keep up with the latest news and updates.
Now, about my trip to Denver … I discovered in a rather funny and ironic way that by asking a chatbot, “Are you sure?”, you can get wildly different answers. Like I asked, “Hey Bard! Can I catch the A Line in Denver from a Bustang stop?”
Well, the answer is no but not before Bard confused the hell outta me.
Anyway, stay tuned for my upcoming Bustang story. Meanwhile, I think I’ll publish Keepin’ Up with the Chatbots on Mondays. Read on, subscribe, share, and thank you for reading!