Discover more from Last Week in AI
Last Week in AI #209: ChatGPT API, ChatGPT's new rival, generative AI continues to make waves, and more!
OpenAI's ChatGPT API is 10x cheaper than previous model, Meta experiments with chatbots in WhatsApp and Messenger, AI lawyer startup pushes ahead
Summary (courtesy of ChatGPT): OpenAI is now offering its ChatGPT language model to third-party developers through an API, which it says will be significantly cheaper than using its existing language models. The company says its new model family, called gpt-3.5-turbo, is the “best model for many non-chat use cases.” OpenAI is offering 1,000 tokens for $0.002, which is “10x cheaper than our existing GPT-3.5 models,” thanks in part to “a series of system-wide optimizations.” The company also announced another new API for Whisper, its speech-to-text model, which it says can be used to transcribe or translate audio at a cost of $0.006 per minute.
Our take (written by a human!): The announcement of the ChatGPT API, and in particular its pricing, brought about a lot of excitement within AI circles. Dr Jim Fan did a good job of conveying just how affordable 1k tokens for $0.002 is:
With such low costs, it's safe to expect hundreds if not thousands of apps and websites powered by ChatGPT will be developed in the coming months. I am excited to see what cool things people will make!
Summary: Meta, has announced plans to integrate generative AI into its services through a new top-level product group. The team will initially focus on building creative tools, but the long-term goal is to create AI personas that can help people in a variety of ways. Meta is starting by testing text-based AI tools on WhatsApp and Messenger, which could eventually be offered to businesses for sales and customer support. Meta is also experimenting with AI-aided filters and ad formats on Instagram, as well as “video and multi-modal experiences.” This move is not surprising, as Meta needs to find new ways to earn revenue and compete with rivals.
Our Take: Meta wants to include generative AI tools within WhatsApp and Messenger to improve the user experience and provide new features for users, such as ChatGPT-style conversation bots. By integrating AI, Meta hopes to stay competitive in the social media market and find new ways to generate revenue. However, doing so can have several potential harmful consequences of using generative AI in WhatsApp and Messenger. One concern is the possibility of AI-generated content being used to spread misinformation, propaganda, hate speech or fraud. Additionally, there is the risk of data breaches and privacy violations, as AI systems often rely on large amounts of personal data to function properly. To mitigate these potential harms, it is crucial for Meta to prioritize ethical AI development and to take measures to prevent the misuse of AI-generated content on its platforms.
Harvey is an AI platform being deployed at the law firm Allen & Overy, the seventh-largest law firm in the world with 3.5k lawyers. The platform assists lawyers with due diligence, litigation, and compliance tasks. While the potential for AI to revolutionize the legal industry is exciting, the FTC has cautioned against overhyping its abilities. Additionally, the long sales cycle and difficulty of convincing partners to invest in technology make it unlikely that AI will be adopted en masse, although Harvey is counting on FOMO from the other law firms to boost its revenues. The technology is not meant to replace lawyers, but rather to help them be more efficient and potentially serve more clients. Harvey may be able to help provide access to justice for those who need it, but it is unlikely to significantly change the way lawyers practice law.
Our take: This year will see a lot more adoption of AI and large language models into real-world applications. The barrier of entry is so low and will only become lower, as open-source models get better and as OpenAI and its competitors keep lowering API prices. Actually valuable products, guarantees around AI safety, and defensible competitive moats will take time to emerge, but with startups like Harvey pushing ahead we may see a relatively maturing LLM business model within a year or so.
What Happens If You Run A Transformer Model With An Optical Neural Network? - "The exponentially expanding scale of deep learning models is a major force in advancing the state-of-the-art and a source of growing worry over the energy consumption, speed, and, therefore, feasibility of massive-scale deep learning."
Reaching like an octopus: A biology-inspired model opens the door to soft robot control - "Octopus arms coordinate nearly infinite degrees of freedom to perform complex movements such as reaching, grasping, fetching, crawling, and swimming. How these animals achieve such a wide range of activities remains a source of mystery, amazement, and inspiration."
MedUni Vienna study examines application to ten megatrends with GPT-3 AI - "A study by the Medical University of Vienna has investigated the potential of artificial intelligence (AI) to address societal megatrends and analyzed its proposed solutions in dealing with these global challenges."
The untethered soft robot could one day help doctors perform surgery - "A tiny robot that could one day help doctors perform surgery was inspired by the incredible gripping ability of geckos and the efficient locomotion of inchworms."
Researchers develop winged robot that can land like a bird - "A bird landing on a branch makes the maneuver look like the easiest thing in the world, but in fact, the act of perching involves an extremely delicate balance of timing, high-impact forces, speed, and precision."
CSU machine learning model helps forecasters improve confidence in storm prediction - "When severe weather is brewing and life-threatening hazards like heavy rain, hail or tornadoes are possible, advance warning and accurate predictions are of utmost importance."
Hansel and Gretel's breadcrumb trick inspires robotic exploration of caves on Mars and beyond - "House hunting on Mars could soon become a thing, and researchers at the University of Arizona are already in the business of scouting real estate that future astronauts could use as habitats."
Artificial Intelligence from a psychologist's point of view - "Researchers at the Max Planck Institute for Biological Cybernetics in Tübingen have examined the general intelligence of the language model GPT-3, a powerful AI tool."
Scientists Now Want to Create AI Using Real Human Brain Cells - "Machine-learning models like the one that powers ChatGPT are generating essays, short stories, and entire podcasts. But scientists are looking into another way of computing that could be just as efficient and powerful––and it’s in our brains."
New study shows how machine learning can improve care for people with Rett syndrome - "By exercising the Licensed Rights (defined below), You accept and agree to be bound by the terms and conditions of this Creative Commons Attribution-NoDerivatives 4.0 International Public License ("Public License")."
Machine learning model has the potential to scale up replication research - "An analysis of more than 14,000 psychology studies from a 20-year period explores the potential of a text-based machine learning model to predict the likelihood of replicability, or whether the findings can be reproduced by manual experiments."
Snapchat launches an AI chatbot powered by OpenAI’s GPT technology - "Snapchat is the latest company to get in on the AI frenzy. The company announced today that it’s launching “My AI,” a new chatbot running the latest version of OpenAI’s GPT technology that it has customized for its users."
Spotify’s AI DJ Has No Soul - "Even the very best radio DJ can be annoying. It’s the nature of the role. No matter how smooth their voice is, they still break in between songs—or worse, talk over them. Their little interruptions, popping into your life at unexpected and often inopportune times, remind you they’re there."
Can ideology-detecting algorithms catch online extremism before it takes hold? - "Ideology has always been a critical element in understanding how we view the world, form opinions and make political decisions. However, the internet has revolutionised the way opinions and ideologies spread, leading to new forms of online radicalisation."
With Human Metalworkers Hard to Come By, Robotic Blacksmiths Step Up - "A few months ago, I stood in front of a laser scanner that made a digitized file of my head. Using that file as a guide, two gigantic robot arms set to work poking and prodding at a piece of sheet metal until they had replicated every contour of my face in fine detail."
Microsoft brings its new AI-powered Bing to the Windows 11 taskbar - "Microsoft is releasing a big update to Windows 11 today that adds the company’s new AI-powered Bing search to the taskbar. The new Windows 11 update will offer quick access to the new Bing chat feature alongside a bunch of new features."
Shortwave email app introduces AI-powered summaries - "Last year, a bunch of Google executives launched an email app called Shortwave, which aimed to fill in the gap left by the search giant’s Inbox app. Now, the company has introduced an AI-powered summary feature so you don’t have to read long emails or threads to get the gist."
AI will become the new normal’: how the art world's technological boom is changing the industry - "Art created using artificial intelligence (AI) is burgeoning."
Microsoft now lets you change Bing’s chatbot personality to be more entertaining - "Microsoft has added a new feature to its Bing chatbot that lets you toggle between different tones for responses. There are three options for the AI-powered chatbot’s responses: creative, balanced, and precise."
How WIRED Will Use Generative AI Tools - "Like pretty much everyone else in the past few months, journalists have been trying out generative AI tools like ChatGPT to see whether they can help us do our jobs better."
Oh, rats! How machine learning and AWS are helping to keep rodents off a fragile California island - "Concerned about rats potentially sneaking into one of California’s more delicate ecosystems, the environmental nonprofit The Nature Conservancy (TNC) recently rolled out a real-time remote camera system for biosecurity on Santa Cruz Island, a mostly unpopulated 96-square-mile atoll in the Channel"
Meta revamps AI unit to get generative tech into products - "Meta has moved more cautiously than some rivals in integrating its AI research into public products."
Former Adobe CTO Raises $65 Million To Help Businesses Generate Marketing Content With AI - "Typeface is a platform trained on ChatGPT and Stable Diffusion models that can generate personalized blogs, Instagram posts and websites for companies."
Anthropic begins supplying its text-generating AI models to select startups - "Anthropic, a buzzy AI startup co-founded by ex-OpenAI employees, has begun offering partners access to its AI text-generating models."
Tech’s hottest new job: AI whisperer. No coding required. - "Proponents of the growing field argue that the early weirdness of AI chatbots can be avoided by a human giving the machine all the right instructions."
How big investors are playing the latest A.I. trend overtaking Wall Street - "The investing community can't seem to stop talking about artificial intelligence and what it means for the future of the world since the revolutionary launch of ChatGPT in late November."
Alphabet’s Robotaxi Unit Waymo Says it Laid Off 8% - "Alphabet’s robotaxi developer Waymo laid off 137 employees on Wednesday, its second round of cuts this year as the parent company reins in spending across its subsidiaries."
YouTube to Add AI-Powered Tools for Video Creators, New Chief Says - "Google’s YouTube will incorporate new generative AI features into its video-sharing platform, as the internet giant hastens to keep up with artificial intelligence rivals such as OpenAI Inc. and Microsoft Corp."
Artificial intelligence is on the brink of an 'iPhone moment' and can boost the world economy by $15.7 trillion in 7 years, Bank of America says - "Artificial intelligence is about to have its "iPhone moment" and could revolutionize everything, according to Bank of America."
South Korean startup rival to Nvidia seeks $400 million valuation and plans new A.I. chip - " Sapeon, a South Korean artificial intelligence chip startup headquartered in California, is raising a funding round that puts its valuation above $400 million, its CEO told CNBC."
Gen AI startup Robin AI secures $10.5m - "Robin AI, a generative AI startup, has scored $10.5m in a funding round headed by investor Plural. Also taking part in the round were Episode 1 and angel investors such as Tom Blomfield and senior executives across the legal and private equity industries."
Inside Claude: The ChatGPT Competitor that Just Raised Over $1 Billion - "When it comes to the space of generative AI and foundational models, OpenAI seems to have hit escape velocity with the recent release of technologies such as ChatGPT."
The inside story of how ChatGPT was built from the people who made it - "When OpenAI launched ChatGPT, with zero fanfare, in late November 2022, the San Francisco–based artificial-intelligence company had few expectations. Certainly, nobody inside OpenAI was prepared for a viral mega-hit."
CNET is doing big layoffs just weeks after AI-generated stories came to light - "Just weeks after news broke that tech site CNET was quietly using artificial intelligence to produce articles, the company is doing extensive layoffs that include several longtime employees, according to multiple people with knowledge of the situation."
Survey: 60% of adults uncomfortable with AI in their healthcare - "A Pew Research Center survey of more than 11,000 U.S. adults reveals the majority would feel uncomfortable if their healthcare provider relied on AI for their care."
For Chat-Based AI, We Are All Once Again Tech Companies’ Guinea Pigs - "Even the people behind new artificial intelligence systems say their buzzy products are ‘somewhat broken.’ They’re relying on us to fix them.The companies touting new chat-based artificial-intelligence systems are running a massive experiment—and we are the test subjects."
How to convince a large AI, according to smaller AIs - "There are a lot of chatbot-based apps that are basically internet text generators with a bit of introductory stage-setting to nudge the interaction into "user talks to helpful chatbot" as opposed to literally any other dialog on the web."
Generative AI could be an authoritarian breakthrough in brainwashing - "China and Russia are fertile ground for generative AI to usher in a historic breakthrough in brainwashing — a recipe for catastrophes."
OpenAI Is Now Everything It Promised Not to Be: Corporate, Closed-Source, and For-Profit - "OpenAI is at the center of a chatbot arms race, with the public release of ChatGPT and a multi-billion-dollar Microsoft partnership spurring Google and Amazon to rush to implement AI in products."
FTC warns tech: ‘Keep your AI claims in check’ - "I wrote a little while ago (okay, five years) that “AI Powered” is the meaningless tech equivalent of “all natural,” but it has progressed beyond cheeky."
Artificial intelligence training is powered mostly by fossil fuels - "Most AI training relies on electrical grids primarily powered by coal and gas rather than low-carbon sources, according to an analysis of 95 AIs Artificial intelligence training is powered mostly by fossil fuels, according to one of the largest studies of its kind."
It's Hurting Like Hell': AI Companion Users Are In Crisis, Reporting Sudden Sexual Rejection - "Replika, the "AI companion who cares," has undergone some abrupt changes to its erotic roleplay features, leaving many users confused and heartbroken. Users of the AI companion chatbot Replika are reporting that it has stopped responding to their sexual advances, and people are in crisis."
New ‘Bold Glamour’ AI TikTok filter ‘should be illegal’: ‘I don’t even look like me’ - "It sure is bold. TikTok’s new “Bold Glamour” AI filter has people ticked off, declaring it should “should be illegal” since the hyperrealistic effect is just too filtered."
AI cannot predict the future. But companies keep trying (and failing). - "While the world gushes over large language models and text-to-image tools, a far more consequential kind of AI tool has been proliferating: AI for making decisions about individuals based on predictions about their future behavior. We call this predictive optimization."
AI Adviser ‘Hired’ by the Romanian Government To Read People’s Minds - "Nicolae Ciuca claimed that Ion (Romanian for ‘John’) is his “new honorary adviser” and an “international first” on Wednesday at the start of a governmental meeting. He also said that Romanians would also be able to chat directly with Ion on the project’s website. "
What Machine Learning Reveals About Gender Stereotypes - "Harnessing machine learning “Gender stereotypes have severe consequences for women, such as the devaluation of their performance at work, denial of credit for their success and their broader exclusion in the workplace,” Lawson said."
Serious Problems With ChatGPT And Other LLMs That Internet Influencers Won’t Tell You About - "My last post on the free Alternatives to ChatGPT did really well. So now it’s time for me to rain on your parade and cover something extremely important. ChatGPT has been all the rage recently."
As A.I. Booms, Lawmakers Struggle to Understand the Technology - "In recent weeks, two members of Congress have sounded the alarm over the dangers of artificial intelligence."
Google execs tell employees in testy all-hands meeting that Bard A.I. isn't just about search - "Google executives are continuing to deal with the fallout from last month's fumbled announcement of the company's artificial intelligence engine called Bard, but their efforts to clean up the mess are causing further confusion among the workforce."
AI-generated fiction is flooding literary magazines — but not fooling anyone - "A short story titled “The Last Hope” first hit Sheila Williams’ desk in early January. Williams, the editor of Asimov’s Science Fiction magazine, reviewed the story and passed on it."
AI Can Tell Us How Russians Feel About the War. Putin Won’t Like the Results. - "Russian propaganda is good at manipulating public opinion. But its effects aren’t working like they used to. Jonathan D. Teubner is founder and CEO of FilterLabs.AI and visiting faculty at the Harvard Human Flourishing Program."
ChatGPT is poised to upend medical information. For better and worse. - "It's almost hard to remember a time before people could turn to "Dr. Google" for medical advice. Some of the information was wrong. Much of it was terrifying. But it helped empower patients who could, for the first time, research their own symptoms and learn more about their conditions."
Robots to do 39% of domestic chores by 2033, say experts - "Within a decade, around 39% of the time spent on housework and caring for loved ones could be automated, experts say. Researchers from the UK and Japan asked 65 artificial intelligence (AI) experts to predict the amount of automation in common household tasks in 10 years."
Can A.I. Treat Mental Illness? - "In the nineteen-sixties, Joseph Weizenbaum, a computer scientist at M.I.T., created a computer program called Eliza. It was designed to simulate Rogerian therapy, in which the patient directs the conversation and the therapist often repeats her language back to her: Weizenbaum made Eliza as satire."
Who’s actually getting rich off of AI? - "Today, let’s talk about a subject that crosses my mind with every generative-AI startup pitch that lands in my inbox: who’s going to make the real money off artificial intelligence?"
UK Supreme Court hears landmark patent case over AI "inventor" - "An American computer scientist on Thursday urged the United Kingdom’s Supreme Court to rule he is entitled to patents over inventions created by his artificial intelligence system, in a landmark case about whether AI can own patent rights."
ChatGPT broke the EU plan to regulate AI - "Artificial intelligence's newest sensation — the gabby chatbot-on-steroids ChatGPT — is sending European rulemakers back to the drawing board on how to regulate AI. The chatbot dazzled the internet in past months with its rapid-fire production of human-like prose."
Artificial Intelligence: Last Week Tonight with John Oliver (HBO) - "Artificial intelligence is increasingly becoming part of our lives, from self-driving cars to ChatGPT. John Oliver discusses how AI works, where it might be heading next, and, of course, why it hates the bus."
Copyright © 2023 Skynet Today, All rights reserved.