

Discover more from Last Week in AI
Last Week in AI #206: Microsoft adds ChatGPT to Bing, responses from Google, Meta, Alibaba, and more!
Microsoft announces major push in AI with OpenAI's help, other tech giants are forced to respond
Top News
A Tech Race Begins as Microsoft Adds A.I. to Its Search Engine
Amidst the tech industry’s worst slump in decades, Microsoft and OpenAI have patterned to put a ChatGPT-like interface to Microsoft’s Bing search engine and Edge browser. The company is hoping that the introduction of ChatGPT will help them catch up to competitors and become the first big company to tech’s next big thing, unlike how it missed the trends on smartphones. The new Bing search engine was demonstrated at a press event and is capable of planning a vacation, researching Japanese poets, translating results, and more. While currently still in beta, the new Bing is expected to be rolled out to millions of people by the end of the month. Other companies such as Google, Meta, and various start-ups are also jumping into the chatbot race. However, there is a concern that the technology may not live up to expectations, as has been the case with other new technologies in the past. Microsoft and OpenAI are aware of the potential downsides and are calling for conversations about the issues.
Our take: Last week’s contrast between Microsoft and Google is stark - while Microsoft was quickly pushing ChatGPT to Bing, Google’s demo of Bard, a similar chatbot search service, was seen as rushed and ill-prepared. While ad sales make up an insignificant portion of Microsoft’s margins, the story is very different for Google. This together with the reputation of having a more popular and perhaps reliable search engine is making Google scale these AI services much slower than its competitors. Credits to Microsoft for “forcing” the other tech giant’s hand, so to speak, and speeding up the deployment of Large Language Models.
Meta, Long an A.I. Leader, Tries Not to Be Left Out of the Boom
As Google and Microsoft attempt to catch up to new industry disruptors like OpenAI, Meta appears to be left behind in the generative AI race, prompting questions about the reason for its shortcomings and its path moving forward. Meta has launched AI products in the past: the company’s Galactica system, a chatbot designed to aid in scientific research, was released two weeks before ChatGPT. Yet it was pulled from the internet after only three days as major issues with fake information caused it to generate widespread criticism. BlenderBot, another chatbot launched by Meta last year, faced the opposite issue: as a result of its many safeguards against problematic content, users found the system overly sanitized and uninteresting. Meta’s past reputation of spreading disinformation and misusing personal data now appears to be stalling its progress in generative AI, as its actions are more heavily criticized than those of smaller companies. Despite struggling to live up to its “move fast and break things” motto of the past, Meta is still investing serious money into AI, and Zuckerberg has been hinting at new AI products and services. These products are likely to build upon Meta’s existing AI systems used for targeting ads, filtering content, and recommending posts to users. As Meta struggles to rebrand itself in the rapidly growing market for generative AI, users and businesses wait to see the specific direction these projects will take.
Our take: While Meta fails to be a major player in the recent AI competition, its shortcomings provide valuable insight into the role of public perception in AI development. On one hand, it demonstrates that big tech companies can only lose so much public trust: while recent lawsuits have emerged against Meta for a variety of practices, others have largely evaded government action. These developments may indicate that while the mistakes of big tech platforms may not cost them their short-term user base, they could stall attempts at innovation later on. On the other hand, this raises the question of if people are too uncritically accepting of technologies like ChatGPT when they come from newer companies. While the right balance between caution and innovation is difficult to find, it’s likely that Meta will continue to fall behind in the AI race without a major change. For one, the existing services Zuckerberg mentions expanding aren’t groundbreaking: using AI to produce better-targeted ads and more personalized content doesn’t offer much value to the public, and reinforces the existing reputation of Meta as a profit-seeking corporate giant. Furthermore, the Metaverse has continued to underperform expectations, and with open demos for generative AI coming at a hefty cost, it will be difficult to justify integrating expensive AI into a technology that’s already losing billions.
Alibaba tests ChatGPT rival as Chinese tech giants like Baidu race to build country’s best AI chatbot
Soon after the release of GPT-3, China and a number of other countries released their own foundation models—I discussed how these releases played into an AI nationalist trend in a previous piece. As ChatGPT has captured the public imagination even more than GPT-3 initially did, there's bound to be overseas competition. Indeed, Alibaba Group Holding has said it is testing a ChatGPT-like service. Joining Baidu, which expects to complete testing of the ChatGPT-like system Ernie Bot in March, Alibaba hopes to provide China's answer to OpenAI's chatbot system. Baidu will incorporate Ernie Bot into its search engine, while other Chinese tech giants including Tencent Holdings and JD.com are laying out their own strategies and responses.
Our take: ChatGPT has done much more than GPT-3 to bring AI systems into the public limelight. As I wrote in the AI Nationalism piece mentioned above, there is no doubt that nations across the world will want to develop comparable systems to assert their own positions as AI leaders. But while the previous "arms race" was all about scale—GPT-3 was soon outsized—I think this one will begin to concern interfaces and how "easy" these systems are to interact with. Given the recent semiconductor restrictions, I expect some difficulty in China's ability to compete at a massive scale, at least in the near term. As for other innovations, they and others will come up with, we'll just have to wait and see.
Other News
Research
Machines Learn Better if We Teach Them the Basics - "Imagine that your neighbor calls to ask a favor: Could you please feed their pet rabbit some carrot slices? Easy enough, you’d think. You can imagine their kitchen, even if you’ve never been there — carrots in a fridge, a drawer holding various knives."
Scaling laws for single-agent reinforcement learning - "Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them."
Scaling Laws Literature Review - "Executive summary Scaling laws are predictable relations between the scale of a mode and performance or other useful properties. I have collected a database of scaling laws for different tasks and architectures, and reviewed dozens of papers in the scaling law literature."
AI-Powered FRIDA robot collaborates with humans to create art - "Carnegie Mellon University's Robotics Institute has a new artist-in-residence. FRIDA, a robotic arm with a paintbrush taped to it, uses artificial intelligence to collaborate with humans on works of art. Ask FRIDA to paint a picture, and it gets to work putting brush to canvas."
AI can predict the effectiveness of breast cancer chemotherapy - "Engineers at the University of Waterloo have developed artificial intelligence (AI) technology to predict if women with breast cancer would benefit from chemotherapy prior to surgery. The new AI algorithm, part of the open-source Cancer-Net initiative led by Dr."
Researchers focus AI on finding exoplanets - "New research from the University of Georgia reveals that artificial intelligence can be used to find planets outside of our solar system."
Researchers Discover a More Flexible Approach to Machine Learning - "“Liquid” neural nets, based on a worm’s nervous system, can transform their underlying algorithms on the fly, giving them unprecedented speed and adaptability."
A Study on the Use of Machine Learning Techniques to Predict the Risk of Diabetes - "The following is a summary of “A survey on diabetes risk prediction using machine learning approaches,” published in the November 2022 issue of Primary care by Firdous, et al. Chronic diabetes mellitus (DM) can have a number of negative effects."
Applications
Text-to-audio generation is here. One of the next big AI disruptions could be in the music industry - "The past few years have seen an explosion in applications of artificial intelligence to creative fields. A new generation of image and text generators is delivering impressive results. Now AI has also found applications in music, too."
Machine learning to improve frequent emergency department use prediction: a retrospective cohort study - "Frequent emergency department use is associated with many adverse events, such as increased risk for hospitalization and mortality. Frequent users have complex needs and associated factors are commonly evaluated using logistic regression."
Sports Illustrated’s publisher is using AI to generate fitness advice - "Another major publisher is using AI to generate stories in the name of efficiency."
This New AI Technology Will Turn Your Ex Into a Red Flag, Snake or Even a Dog - "Picsart's "Replace My Ex" might be the next-best solution for the heartbroken."
An important next step on our AI journey - "AI is the most profound technology we are working on today. Whether it’s helping doctors detect diseases earlier or enabling people to access information in their own language, AI helps people, businesses and communities unlock their potential."
How Scientists Are Using AI to Talk to Animals - "In the 1970s a young gorilla known as Koko drew worldwide attention with her ability to use human sign language."
Quora opens its new AI chatbot app Poe to the general public - "Q&A platform Quora has opened up public access to its new AI chatbot app, Poe, which lets users ask questions and get answers from a range of AI chatbots, including those from ChatGPT maker, OpenAI, and other companies like Anthropic."
David Guetta Used AI to Deepfake Eminem Vocals for New Song - "Perhaps unable to land a proper collaboration with Eminem on his own, David Guetta has taken matters into his own hands with the help of AI."
Google’s AI-powered ‘multisearch,’ which combines text and images in a single query, goes global - "Amid other A.I.-focused announcements, Google today shared that its newer “multisearch” feature would now be available to global users on mobile devices, anywhere that Google Lens is already available."
Engineers trained an algorithm to annotate football game film -- a painstaking task currently done manually - "Players and coaches for the Philadelphia Eagles and Kansas City Chiefs will spend hours and hours in film rooms this week in preparation for the Super Bowl."
Business
The race of the AI labs heats up - "Every so often a technology captures the world’s imagination. The latest example, judging by the chatter in Silicon Valley, on Wall Street, in corner offices, newsrooms and classrooms around the world, is ChatGPT."
Analysis: From Meta to Microsoft, AI's big moment is here - "Feb 3 (Reuters) - Big Tech companies have a new obsession: artificial intelligence. This week, chief executives across the sector packed earnings calls with mentions of the heavily hyped technology, which until recently existed more in the background than as a solid contributor to the bottom line."
Hustle bros are jumping on the AI bandwagon - "The grind-set, side-hustle, passive-income crew has a new favorite toy: ChatGPT."
Apple CEO Tim Cook says AI will eventually 'affect every product and service we have' - "Recent advancements in artificial intelligence have caught much of the business community's attention, and Apple CEO Tim Cook is no exception."
The Creator of ChatGPT Thinks AI Should Be Regulated - "In 2023, TIME will once again recognize 100 businesses making an extraordinary impact around the world. Applications for the TIME100 Most Influential Companies of 2023 are open, now through March 1, 2023. Apply here."
AI startup Cohere in talks to raise funding at $6 bln plus valuation - "Feb 6 (Reuters) - Cohere, an AI foundation model company that competes with Microsoft-backed OpenAI, is in talks to raise hundreds of millions of dollars in a funding round that could value the startup at more than $6 billion, sources told Reuters, in the latest sign of the investment frenzy around"
Moonhub wants to transform the way companies find job candidates using AI - "Moonhub founder and CEO Nancy Xu was studying for her computer science PhD at Stanford when she had an epiphany about hiring."
A.I. will be the most important technology shaping the future of trading - JPMorgan survey - "Editor's Note: With so much market volatility, stay on top of daily news! Get caught up in minutes with our speedy summary of today's must-read news and expert opinions. Sign up here!"
Tesla’s former head of AI joins Elon Musk-founded OpenAI - "Tesla’s former head of AI, Andrej Karpathy, announced that he is joining OpenAI, an AI startup originally co-founded by Tesla CEO Elon Musk. Karpathy joined Tesla more than five years ago as a neural net and computer vision expert."
Waymo provides driverless rides for Super Bowl visitors - "Uber and Lyft drivers are busy this week with all the visitors in town for the Super Bowl and the WM Phoenix Open. But there’s another option if they need a ride around Phoenix. Waymo has been offering driverless rides around the city for a few years."
Concerns
How ChatGPT Kicked Off an A.I. Arms Race - "One day in mid-November, workers at OpenAI got an unexpected assignment: Release a chatbot, fast. The chatbot, an executive announced, would be known as “Chat with GPT-3.5,” and it would be made available free to the public. In two weeks."
Scams, deepfake porn and romance bots: advanced AI is exciting, but incredibly dangerous in criminals’ hands - "The generative AI industry will be worth about A$22 trillion by 2030, according to the CSIRO. These systems – of which ChatGPT is currently the best known – can write essays and code, generate music and artwork, and have entire conversations."
OpenAI's Whisper is another case study in Colonisation - "On 21 September OpenAI dropped Whisper, a speech recognition model trained on 680,000 hours of audio taken from the web. The highlight: The ability for a single model to transcribe in multiple languages is ground-breaking for natural language processing (NLP) technologies."
AI ‘Seinfeld’ show suspended by Twitch for transphobic, homophobic stand-up - "An AI-generated “Seinfeld” parody, which has been streaming for months on Twitch, is now suspended after the computer rendering of Jerry Seinfeld launched into a transphobic stand-up comedy routine."
Getty Images Sues Stability AI Over Art Generator IP Violations - "Visual media company Getty Images accused Stability AI Inc. of using more than 12 million Getty photos without permission or compensation—the latest copyright and trademark lawsuit against the use of artificial intelligence tools to generate art. The complaint, filed Feb."
Conservatives Are Obsessed With Getting ChatGPT to Say the N-Word - "More specifically, they are constructing elaborate scenarios to try and trick OpenAI's ChatGPT tool—which is not sentient, does not understand anything, and is really just a more convincing version of SmarterChild—into saying the racist slur."
ChatGPT is a data privacy nightmare. If you’ve ever posted online, you ought to be concerned - "ChatGPT has taken the world by storm. Within two months of its release it reached 100 million active users, making it the fastest-growing consumer application ever launched."
I asked Microsoft's 'new Bing' to write me a cover letter for a job. It refused, saying this would be 'unethical' and 'unfair to other applicants.' - "The rise of artificial intelligence powered chatbots such as ChatGPT have raised questions about the ethics of using such platforms for job searches and exams, but Microsoft's new AI-enabled Bing search engine seems to have certain built-in controls."
Developers Created AI to Generate Police Sketches. Experts Are Horrified - "The program, called Forensic Sketch AI-rtist, was created by developers Artur Fortunato and Filipe Reynaud as part of a hackathon in December 2022."
ChatGPT embraced by hackers, but some AI experts say it’s not botmageddon—yet - "There is no technology that can’t, and won’t, be used for both good and evil. So, no surprise, that’s what’s happening with ChatGPT, the latest, hottest tech toy."
Could ChatGPT supercharge false narratives? - "Many warn of the tool's potential to be a misinformation superspreader, capable of instantly producing news articles, blogs and political speeches. Brady is expected to head to the booth as the highest-paid TV analyst in the history of the game … even though he has never announced before."
Google’s AI chatbot Bard makes factual error in first demo - "On Monday, Google announced its AI chatbot Bard — a rival to OpenAI’s ChatGPT that’s due to become “more widely available to the public in the coming weeks.” But the bot isn’t off to a great start, with experts noting that Bard made a factual error in its very first demo."
AI technologies have even more exaggerated biases in perception of facial age than humans - "Artificial intelligence algorithms are the wave of the future."
The current legal cases against generative AI are just the beginning - "As generative AI enters the mainstream, each new day brings a new lawsuit."
ChatGPT Is a Blurry JPEG of the Web - "In 2013, workers at a German construction company noticed something odd about their Xerox photocopier: when they made a copy of the floor plan of a house, the copy differed from the original in a subtle but significant way."
US experts warn AI likely to kill off jobs – and widen wealth inequality - "ChatGPT is just the latest technology to fuel worries that it will wipe out the jobs of millions of workers, whether advertising copywriters, Wall Street traders, salespeople, writers of basic computer code or journalists."
How to Stop AI From Eating Journalism - "All at once, it seems, the many dire warnings about artificial intelligence are coalescing into reality."
Is ChatGPT ‘woke’? AI chatbot accused of anti-conservative bias and a grudge against Trump - "Ask ChatGPT about drag queen story hours or former president Donald Trump and conservatives say it spits out answers that betray a distinct liberal bias. Allegations that ChatGPT has gone “woke” began circulating after a recent National Review article."
Silicon Valley Is Turning Into Its Own Worst Fear - "Because corporations lack insight, we expect the government to provide oversight in the form of regulation, but the internet is almost entirely unregulated."
ChatGPT Can't Pick Between Fairmount & St-Viateur Bagels — Maybe It's Scared Of Montrealers - "For decades, the Great Montreal Bagel Debate has stirred tepid interest among acquaintances at house parties who can't think of anything more interesting to discuss."
AI is the latest Wall Street craze. Is it also the next bubble? - "Artificial intelligence is the latest tech flavor of the month. Industry giants Google, Microsoft and China’s Baidu have all had big AI announcements in recent days, as ChatGPT bot mania is taking the corporate world by storm."
The Generative AI Race Has a Dirty Secret - "In early February, first Google, then Microsoft, announced major overhauls to their search engines. Both tech giants have spent big on building or buying generative AI tools, which use large language models to understand and respond to complex questions."
Is A.I. Art Stealing from Artists? - "Last year, a Tennessee-based artist named Kelly McKernan noticed that their name was being used with increasing frequency in A.I.-driven image generation. McKernan makes paintings that often feature nymphlike female figures in an acid-colored style that blends Art Nouveau and science fiction."
Analysis
Inside the Heart of ChatGPT’s Darkness - "In hindsight, ChatGPT may come to be seen as the greatest publicity stunt in AI history, an intoxicating glimpse at a future that may actually take years to realize—kind of like a 2012-vintage driverless car demo, but this time with a foretaste"
Policy
The accelerating use of generative AI may prompt U.S. action - "Ammanath said there's "no question" that AI regulations are coming. Efforts are underway to understand better how AI works, its impact on consumers, and how to hold AI developers and users accountable for fair and transparent systems. The U.S."
Italy bans U.S.-based AI chatbot Replika from using personal data - "MILAN/LONDON (Reuters) - Italy's Data Protection Agency said on Friday it was prohibiting artificial intelligence (AI) chatbot company Replika from using the personal data of Italian users, citing risks to minors and emotionally fragile people."
Why Chinese AI and semiconductors could fall decades behind under US chip ban ‘blitz’ - "As the US pushes to keep China from getting the latest chip-making technology, tightened sanctions could cause Chinese semiconductor and AI development to fall decades behind, according to industry insiders."
NIST Risk Management Framework Aims to Improve Trustworthiness of Artificial Intelligence - "The U.S. Department of Commerce’s National Institute of Standards and Technology (NIST) has released its Artificial Intelligence Risk Management Framework (AI RMF 1."
U.S. Copyright Office tells Judge that AI Artwork isn’t Protectable - "The U.S. Copyright Office has told a federal judge that artificial intelligence (AI) artwork can’t be protected. The Copyright Office is attempting to get a lawsuit brought against them by Stephen Thaler dismissed."
Expert Opinions
In the Age of A.I., Major in Being Human - "Last summer, a piece of artwork generated with artificial intelligence took a first prize at the Colorado State Fair. To me, the image looks like a view from the back of the stage at an opera."
Explainers
Understanding the Self-Attention Mechanism of Large Language Models From Scratch - "In this article, we are going to understand how self-attention works from scratch. This means we will code it ourselves one step at a time."
Fun
Artificial Intelligence Just Lost a Leader - "Roger Schank just passed away. Roger was a top leader of AI. I overlapped with him for my time at Yale. In 1974, he became a professor of computer science and psychology at Yale University."
My Strange Day With Bing’s New AI Chatbot - "Twenty minutes after Microsoft granted me access to a limited preview of its new chatbot interface for the Bing search engine, I asked it something you generally don’t bring up with someone you just met: Was the 2020 presidential election stolen?"
Copyright © 2023 Skynet Today, All rights reserved.
Last Week in AI #206: Microsoft adds ChatGPT to Bing, responses from Google, Meta, Alibaba, and more!
Great article man! Very comprehensive and to the point. How do you go about collecting such information to write about? It is quite remarkable and would like to crawl around the way you have done for this article.
There's really *a lot* going on!