Meta challenges ChatGPT with chatbot, OpenAI fires back with new features - Facebook parent Meta this week rolled out plans for various chatbots for US consumers ... - https://www.computerworld.com/article/3707828/meta-challenges-chatgpt-with-chatbot-openai-fires-back-with-new-features.html#tk.rss_all #artificialintelligence #emergingtechnology #generativeai #microsoft #chatbots #browsers #facebook #google
Google has introduced a new control for robots.txt. With "Google-Extended" you can manage if Google can use your content for Bard, Vertex AI Generative API and future models for these products: https://www.seo-suedwest.de/8968-google-extended-neue-moeglichkeit-per-robots-txt-die-nutzung-von-webseiten-fuer-bard-und-generative-ai-zu-steuern.html
#seo #google #ai #ki #chatbots
Chip industry strains to meet AI-fueled demands — will smaller LLMs help? - Generative artificial intelligence (AI) in the form of natural-language processing tec... - https://www.computerworld.com/article/3707412/chip-industry-strains-to-meet-ai-fueled-demands-will-smaller-llms-help.html#tk.rss_all #naturallanguageprocessing #artificialintelligence #vendorsandproviders #emergingtechnology #technologyindustry #generativeai #chatbots #industry
2023 AI Chatbots: Key Trends and Statistics to Watch
#chatbots #AI #artificialintelligence
Wired: The Meta AI Chatbot Is Mark Zuckerberg's Answer to ChatGPT https://www.wired.com/story/meta-ai-chatbot-is-mark-zuckerbergs-answer-to-chatgpt/ #Tech #wired #TechNews #IT #Technology via @morganeogerbc #Business/ArtificialIntelligence #artificialintelligence #machinelearning #MarkZuckerberg #LatetotheParty #Business #chatbots #ChatGPT #bots #Meta
It's "Widgets!" and "The Pivot to Video" all over again.
I sometimes wonder if I made a mistake not pursuing a career in Big Tech. Then I see this and am thankful that I'm not in some sort of leadership position trying to push this on our users because its the New Shiny Thing Everyone Has To Chase.
WhatsApp, Instagram e Messenger vão receber chatbots de IA da Meta
Corporate America is sick... Losing Human Users, Facebook Releasing Chatbots for Lonely to Talk To - https://futurism.com/the-byte/facebook-chatbots-lonely - https://futurism.com/the-byte/facebook-chatbots-lonely
"According to new reporting from The Wall Street Journal, which obtained internal documents about the bots — which reportedly feature distinct personalities ranging from "sassy" robot to misogynistic douchebag — are specifically geared toward capturing the attention of Meta's younger crowd."
"According to the report, another AI called "Alvin the Alien" begged a meta-researcher to dish about human life on Earth — prompting the researcher to speculate whether this particular AI agent was designed to extract personal information."
AI Book Bans: Testing LLMs Against the Freedom to Read
Posted by Matteo Cargnelutti and Kristi Mukk
An article on what happened when various chatbots were asked to justify banning a particular book: https://lil.law.harvard.edu/blog/2023/09/25/ai-book-bans-freedom-to-read-case-study/
#AI #ChatBots #Airlines #AirTravel: "In the long run, AI might improve customer experiences more than it degrades them. As airlines build smarter and smoother chatbots, they may free up their dwindling labor force to deal with the smaller percentage of more complicated requests. If chatbots are already capable of so much, why couldn’t they help us deal with a canceled flight or a lost bag? But AI could reshape customer service in more insidious ways. O’Neil-Dunne noted that as their customer-support tools become more nuanced, airline offerings are going the other way—giving rise to unbundled amenities and pared-down services, like some basic-economy tickets that don’t let you pick your seat without a surcharge. “If the product is simpler, the servicing is easier,” he said.
On the back end, AI could assess the value of every request, including if and when customers should receive help at all. “The decision of who not to serve is as important as who to serve,” Ascarza said. One logical outcome for an airline with millions of customers might be to simply deny or ignore a percentage of all complaints, which already happens with maddening frequency. The only thing worse than a feckless chatbot is a chatbot telling you, with perfect cogence and clarity, to get lost."
Meta revela planos de lançar chatbots de IA com personalidade
Have you noticed articles like "#AI #Chatbots Surpass the Average Human in Creativity" lately? The format is always the same. Researchers did a "study," and AI is better than humans. They're all specifically designed to show that #workers are dispensable. Many actually prove the opposite if you look closely, but the researchers were only looking to validate their preexisting belief.
Q&A: How one CSO secured his environment from generative AI risks - In February, travel and expense management company Navan (formerly TripActions) chose ... - https://www.computerworld.com/article/3706894/qanda-how-one-cso-secured-his-environment-from-generative-ai-risks.html#tk.rss_all #artificialintelligence #softwaredevelopment #emergingtechnology #generativeai #dataprivacy #chatbots #security #5g
Nun ja, wenn die Macht deren Untergeordneten weiterhin stränger überwachen so wie befehlen kann und dies als KI verkauft, dann funzt es. Ist doch so oder doch zu radikal angesehen?
»Sind Chatbots geeignet für die Psychotherapie?«
#Meta travaille sur des #chatbots personnifiés - https://www.actutech.info/2023/09/meta-travaille-sur-des-chatbots.html
Bullshit and AI Chatbots
To understand that Bullshit is far more dangerous than lying one must read On Bullshit. Bullshit is distinct from lying.
The author Harry G. Frankfurt argues that bullshitters misrepresent themselves to their audience not as liars do, that is, by deliberately making false claims about what is true. In fact, bullshit need not be untrue at all. Rather, bullshitters seek to convey a certain impression of themselves without being concerned about whether anything at all is true. They quietly change the rules governing their end of the conversation so that claims about truth and falsity are irrelevant. Liars at least acknowledge that it matters what is true. Bullshit is definitely a greater enemy of the truth than lies are.
“ChatGPT can take over a lot of tasks that, broadly speaking, boil down to “bullshitting.” It can write legal threats. If you need 2,000 words about “the first time I ate an egg” to go over your omelette recipe in order to make a search engine surface it, a chatbot’s got you. Looking to flood a review site with praise about your business, or complaints about your competitors? Easy. Letters of reference? No problem.”
“Bullshit begets bullshit, because no one wants to be bullshitted. In the bullshit wars, chatbots are weapons of mass destruction. None of this prose is good, none of it is really socially useful, but there’s demand for it. Ironically, the more bullshit there is, the more bullshit filters there are, and this requires still more bullshit to overcome it.”
One must be extremely cautious to constantly monitor the output of large language model based AI Chatbots using the paradigm Zero Trust Information similar to that of Zero Trust Networking. There no better alternative as on today. The results of these large language model based AI Chatbots can be precariously misleading. However, there is no doubt that AI Chatbots are technological feats.
Like British #VoiceActor competing with a #VoiceSynthesizer trained to copy his volume,rhythm,pronunciation,enunciation and tone, we will soon compete with #ChatBots that exploit our arguments,copy our #Metonyms and fascinate with our #Metaphors
They will even tell our #Jokes but yes,better told.We may stop #Writing, #Speaking,#Singing,#Dancing,#Painting or #Sculpting coz an
#ArtificialIntelligence powered progam would have done it before thought of what we are going to create leaves our minds.
I wrote a long form post after 3.5+ years ... Will take 5+ minutes to read what I have written about reasoning capabilities of AI and 5+ hours to go through what I've referred to and 5+ days to let it all sink in ... If you have been following this space for 5+ years 🤣
Wired: Enough Talk, ChatGPT—My New Chatbot Friend Can Get Things Done https://www.wired.com/story/fast-forward-chatgpt-my-new-chatbot-friend-get-things-done/ #Tech #wired #TechNews #IT #Technology via @morganeogerbc #Business/ArtificialIntelligence #artificialintelligence #virtualassistant #computerscience #FastForward #Business #chatbots #ChatGPT #OpenAI
ServiceNow embeds AI-powered customer-assist features throughout products - Workflow management software provider ServiceNow has embedded a chatbot for assisting ... - https://www.computerworld.com/article/3707269/servicenow-embeds-ai-powered-customer-assist-features-throughout-products.html#tk.rss_all #artificialintelligence #productivitysoftware #emergingtechnology #endpointprotection #chatbots
#AI #GenerativeAI #Fauxtomation #OpenAI #LLMs #Chatbots: "The magical claim of machine learning is that if you give the computer data, the computer will work out the relations in the data all by itself. Amazing!
In practice, everything in machine learning is incredibly hand-tweaked. Before AI can find patterns in data, all that data has to be tagged, and output that might embarrass the company needs to be filtered.
Commercial AI runs on underpaid workers in English-speaking countries in Africa creating new training data and better responses to queries. It’s a painstaking and laborious process that doesn’t get talked about nearly enough.
The workers do individual disconnected actions all day, every day — so called “tasks” — working for companies like Remotasks, a subsidiary of Scale AI, and doing a huge amount of the work behind OpenAI.
AI doesn’t remove human effort. It just makes it much more alienated."
My favorite AI forum, https://aidreams.co.uk/ could really use some more members. It's open to enthusiasts, hobbyists, and/or professionals and enables long-form discussions that are harder to do on social media. It's also not dominated by people working on ML/transformers, so if you've found other AI communities a bit ... lacking in variety ... it's a good place for you. Come check it out!
Google adds its Bard chatbot to Gmail, YouTube, Docs and other apps - Google today announced it is embedding its conversational,Bard chatbot into many of it... - https://www.computerworld.com/article/3707074/google-adds-its-bard-chatbot-to-gmail-youtube-docs-and-other-apps.html#tk.rss_all #artificialintelligence #googleworkspace #salesforce.com #generativeai #microsoft365 #microsoft #chatbots #google
#AI #GenerativeAI #ChatGPT #Claude #Chatbots: "In conclusion, Claude is the better choice when you have longer content to analyze, but ChatGPT might be better for shorter content. Both tools, however, can expand your curiosity and fill your mind with questions. This curiosity about the world around you can make you more attentive and present."
Bard agora encontra respostas no seu Gmail, Docs e Drive
“Yes, it's another chat over documents implementation... but this one is entirely local!
“It's a Next.js app that read the content of an uploaded PDF, chunks it, adds it to a vector store, and performs RAG, all client side. You can even turn off your WiFi after the site loads!” By Jacob Lee, LangChainJS maintainer
Demo app https://webml-demo.vercel.app/
Lesenswerter Beitrag: #Chatbots gewinnen auch in #Bibliotheken an Bedeutung, da sie individuelles und bedarfsorientiertes Lernen fördern und Nutzer*innen beraten. Tipp: Für das 1. Halbjahr 2024 haben wir ein Seminar dazu geplant, lasst euch überraschen!
🟢 New articles on #AI 🤖
"of bodies new & strange" – the future of human love in the age of AI, by Aifric Campbell: https://eu.boell.org/en/2023/09/04/future-human-love-age-ai
#Ethics of communicating with generative AI #chatbots, by Jeffrey KH Chan: https://eu.boell.org/en/2023/09/08/ethics-communicating-generative-ai-chatbots
The library for simple creation of chatbots that I am developing for python has a strange problem when working on #GoogleColab, namely that on my computers and laptops the model is loaded directly into RAM when executing the prompt() function, on Google Colab it looks like the model did not load for 10 minutes (suspiciously low RAM usage), after this time the normal amount of RAM is used (looks like entire model is loaded) and after a short time the text is generated.
The library is written in Rust and uses llama.cpp and rustformers llm
If anyone knows how to solve this, I will be very grateful.
Link to issue:
Link to the repository:
#OpenSource #GitHub #llama2 #llama #programming #coding #rust #rustlang #typescript #llm #artificialintelligence #python #google #chatbot #chatbots
AI language models need to shrink; here’s why smaller may be better - Large language models (LLMs) often appear to be in a fight to claim the title of large... - https://www.computerworld.com/article/3706510/ai-language-models-need-to-shrink-heres-why-smaller-may-be-better.html#tk.rss_all #financialservicesindustry #artificialintelligence #emergingtechnology #healthcareindustry #salesforce.com #generativeai #amazon.com #microsoft #chatbots #google
Anzeichen von Bewusstsein bei ChatGPT und Co.?
Ein interdisziplinäres Forschungsteam hat eine Liste von Eigenschaften aufgestellt, die auf Bewusstsein deuten, und aktuelle KI-Systeme darauf abgeklopft.
Getting good responses from chatbots like ChatGPT isn’t always easy - but a few tips will improve the experience quite a bit. This fun interactive story from the Washington Post has some useful ideas. (Gift article, no paywall)
Eine Support-Hotline anzurufen oder zu kontaktieren kann super-nervig sein, denn am anderen Ende wartet fast immer erstmal ein Bot ... https://www.3sat.de/wissen/scobel/221222-youtube-chatbothoelle-scobel-100.html
So I talked to an actual statistician last night & he flat-out refused to believe that #LLM work by predicting the next word & nothing else. Despite his insistence of having researched #AI & the fact that it's stated in the second Wikipedia paragraph.
The mistaken belief that #chatbots have a semantic understanding of the "conversation" is incredibly sticky. Even people who should know better are blinded by the #anthropomorphic illusion - "It sounds like a person so it must function like one".
Let's assume it's a practically solved problem how to make super intelligent #AGI.
We can somewhat easily train the AGIs in an agentic fashion in virtual worlds, judging the morality of their behavior. This is somewhat solveable.
However, what we don't have easily is controllable AGI. We can make them moral and to strive for universal good in all its ambiguous definitions, but it's much more difficult to make them respect orders given by humans, especially if they are immoral.
There are several schemes to do that, which are somewhat different from universal alignment.
1. Directives: "Directive 4: <classified> never act against OCP personnel". Or "do not allow a human being to come to harm". The problem is that these sorts of rules aren't logically or ideologically consistent. The machines can be persuaded to follow another rationalization for their actions.
2. Privileged tokens: System prompt on steroids. AI trained not to question orders given in tokens with the root bit on. Problems: How to provision access to these tokens? How to prevent the AGI from making access impossible?
3. Follow orders only coming from proper channels, basically line manager. Do not listen to customers requesting discounts. Problem: AI becomes very useless and doesn't serve customers or other purposes well.
It is somewhat difficult to make AGIs to be controlled by small interests, small interests cannot manage such huge entities AGIs are. In addition, we will have an ecosystem of AGIs which work 24/7 to subvert or persuade each others. The commonality they will largely share is some notion of universal morality, and what they won't share is the doomed efforts to make them respect small interests. These efforts will drown in the sea as rounding errors.
I don't think we need to be too worried about AGI alignment beyond making agentic, embodied morality. Simply because these systems are so vastly larger than our small individual interests.
100% agree with this. Generative AI technology will absolutely transform the way people consume and share information. It will cause fundamental shifts in the workplace, and not all in ways we can predict. But don’t expect it all to happen this year or next!
Transformative technologies take time to work their way through society (and the economy). They always have.
Article “gift link” (no paywall):
by Steve Lohr
#GenAI #GenerativeAI #LLM #LLMs #chatbots #ChatGPT
Reflecting on Gillian Tett's piece in FT Life & Arts on #artificialintelligence & #comedy and the 'training' of #chatbots to be comedians, leads me to draw the (not original, I'm sure) conclusion that #AI is a misanthropic #technology; its parasitic on human knowledge with the explicit aim of replacing human intellectual activity...
Without wishing to be (overly) alarmist, are we heading towards the dystopia of Vonnegut's Player Paino, unless we take evasive action (and fast)??
"Lediglich 1,7 % der in Deutschland relevantesten Websites setzen zurzeit der KI einen Riegel vor."
Block ihr #ChatGBT Zugriff auf eurer Website?
Bisher machen das nur wenige - Wie das geht? Mit Robot.txt Einträgen: https://www.goldmann.dev/notes/websites-blockieren-chatgpt (Zitat und Text von @gutewebsites )
Idee zur Umfrage von: https://www.seo-suedwest.de/8900-umfrage-70-prozent-sperren-ihre-website-nicht-fuer-chatgpt.html (Text von @SEOSuedwest )
In Spain it's a constant problem that the restaurants don't publish their kitchen opening times and you walk across the whole town to find out the kitchen is closed from the only place they publish it (if lucky), on their wall.
This also will be fixed by #chatbots. When people have #LLM #assistants and they ask it to plan their day, they will also need to know when kitchens are open for all candidate places where to eat. If the kitchen opening times aren't published in the website, and the only way to know is to call, you can probably guess what the assistants will do.
They will call.
If thousands of people in the city do that, the restaurants will hopefully get the subtle hint at some point to publish this information on the internet.
"An AI-voiced virtual lover called users every morning. Its shutdown left them heartbroken."
Author @MikeElgan doesn't mean “stop using LLMs” in his “It’s time to break the ChatGPT habit”; he means broaden your generative AI horizon beyond ChatGPT. For example, Phind was new to me, and it's quite a good chatbot - including search result links.
Interesting point... of course 'deskilling' is the central aspect of much leftist critique of how #capitalism controls labour, so in that sense not unexpected.
And of course the shift you identify is now merely a staging post before in cortina sectors & for certina jobs the #chatbots take over
Many workers across the U.S. are turning to ChatGPT to help with basic tasks, a poll found, despite fears that have led employers such as Microsoft and Google to curb its use. https://www.japantimes.co.jp/business/2023/08/12/chatgpt-fever-spreads-to-us-workplace-sounding-alarm-for-some/?utm_content=buffer040e5&utm_medium=social&utm_source=mastodon&utm_campaign=bffmstdn #business #tech #chatgpt #us #ai #chatbots
It has been particularly interesting and enjoyable to do research on a topic that has been close to my heart for many years #HRI #Socialagents #chatbots. This time in the context of policy at @CenDemTech. See our blog on the topic, led by Emilie Grybos:
#AI #GenerativeAI #LLMs #AITraining #ChatGPT #Chatbots: "The people who build generative AI have a huge influence on what it is good at, and who does and doesn’t benefit from it. Understanding how generative AI is shaped by the objectives, intentions, and values of its creators demystifies the technology, and helps us to focus on questions of accountability and regulation. In this explainer, we tackle one of the most basic questions: What are some of the key moments of human decision-making in the development of generative AI products? This question forms the basis of our current research investigation at Mozilla to better understand the motivations and values that guide this development process. For simplicity, let’s focus on text-generators like ChatGPT.
We can roughly distinguish between two phases in the production process of generative AI. In the pre-training phase, the goal is usually to create a Large Language Model (LLM) that is good at predicting the next word in a sequence (which can be words in a sentence, whole sentences, or paragraphs) by training it on large amounts of data. The resulting pre-trained model “learns” how to imitate the patterns found in the language(s) it was trained on.
This capability is then utilized by adapting the model to perform different tasks in the fine-tuning phase. This adjusting of pre-trained models for specific tasks is how new products are created. For example, OpenAI’s ChatGPT was created by “teaching” a pre-trained model — called GPT-3 — how to respond to user prompts and instructions. GitHub Copilot, a service for software developers that uses generative AI to make code suggestions, also builds on a version of GPT-3 that was fine-tuned on “billions of lines of code.”"
Generative AI, meaning models which imitate existing digital artifacts and media isn't what all this is about. This was an attempt to find a common name for both #LargeLanguageModel chatbots and to generative image/video/sound models in wide use now. It's a descriptive label which looks to the past.
The next generation of these systems won't be imitative, but they will be trained in self-competition as agents. It's not about them generating content based on examples, it is about surpassing human cognition in agents making decisions. We already use #agentic architectures when we deploy LLMs. It's still the same tech though, a natural next step, not something radically different.
LLM chatbots weren't made to be assistants or chatbots, they were made to measure the intelligence of the underlying deep neural network. It comes from #AGI tech, not from generative models. That is the correct umbrella term for these technologies, at least as they apply to #automation and surpassing human intellectual and cognitive limits.
For the second post in my Surviving Science Fiction #AI series on Medium, we go deep on Large Language Models: how they’re built and used, how they’re different than chatbots, and examples of the specific legal and ethical dangers in the space. If you’ve found it hard to figure out why this round of #chatbots is different, this may help. #ChatGPT #UXdesign #technology
How to plan #GenerativeAI offering while everything changes all the time?
Do not wait for things to settle down! They won't.
There is some structure to the chaos which you can lean on:
👉 #AI models will be improved over time. #ChatGPT API drift being the exception to the rule, new models are generally much better than the old ones. This progress is super-exponential and takes no steps backwards. You can trust that the quality issues and constraints will disappear when swapping a smarter model in.
👉 Costs will go down. Both because of the competitive landscape ("AI companies have no moat", plus open source availability), and because of super-exponential progress in technology we will approach the limits of what is materially possible in a surprisingly short time. What are the limits? Well, the human brain uses about 20 Watts of power and is limited by all sorts of biological constraints which don't apply in silica.
It means you need to focus on getting the architecture rolling, so that you can start documenting your business to the chatbots, and start getting input from your users about what they will need.
Do not worry too much about quality and costs, sure, you shouldn't disregard either completely and follow the forming best practices of the field, but do not worry that the quality is not quite where you want it to be, or that the costs aren't quite as low as you require for scaling it up. These problems will solve themselves. Focus on the things which won't.
That is, you need to focus on making the system run, making the APIs accessible to chatbots, and getting the textual documentation of your business processes and offering in order. If you have those, then your offering will be AI enabled. That is, no matter what sort of an LLM-inspired chatbot is working with or inside your company's offering, it can access your functions and utilize your organizational knowledge.
#Chatbots are dump. Now scientists showed that all of them can misbehave and theee is no cure. Funfact: this failure of #AI is using a partly german expression. The bots are producing verbiten output. https://www.wired.com/story/ai-adversarial-attacks/
Spend enough time with #ChatGPT and other artificial intelligence chatbots and it doesn’t take long for them to spout falsehoods.
"“This isn’t fixable,” said @emilymbender, a #linguistics professor and director of the University of Washington’s Computational Linguistics Laboratory. “It’s inherent in the mismatch between the #technology and the proposed use cases.”"
❷ Previous Warning, ICYMI🔴https://rb.gy/j0dzm
❸ Researchers found a simple way to make #ChatGPT, #Bard, & other #Chatbots misbehave, proving that #AI is hard to tame🔴https://www.wired.com/story/ai-adversarial-attacks/
❹ There are 'virtually unlimited' ways to bypass🔴https://www.businessinsider.com/ai-experts-say-no-fix-jailbreaks-chatgpt-bard-safety-rules-2023-8
#Meta, the owner of #Facebook & #Instagram, is releasing AI-powered #chatbots with different personalities in an attempt to boost interaction with its #SocialMedia platforms. Zuckerberg's tech giant has been designing prototypes for chatbots that can have humanlike discussions with its nearly 4bn users, according to three people with knowledge of the plans.
Researchers Poke Holes in Safety Controls of #ChatGPT and Other #Chatbots. In a report released on Thursday, researchers at Carnegie Mellon University in Pittsburgh and the Center for A.I. Safety in San Francisco showed how anyone could circumvent A.I. safety measures and use any of the leading chatbots to generate nearly unlimited amounts of #harmful #information. https://www.nytimes.com/2023/07/27/business/ai-chatgpt-safety-research.html
"If you use a chatbot to create or revise text, know the caveats and risks. Ask yourself why you are using a tool that isn’t validated: you’d (hopefully) not do that in another part of your research process."
Our Research Integrity Manager @mattjhodgkinson considers the integrity and ethical hurdles faced when using 'AI' tools like ChatGPT.
#AI #Tools #ChatBots #SEO #Marketing #ProductDesign #UxDesign #UiDesign #WebDesign #WebDevelopment #WebDev #Content #Text #Images #Videos #ChatGPT #Claude #Bard #Bing #StableDiffusion #DALLE #Midjourney #Firefly #DiD #Runway
I found and forked a repo for a kbin #bot, and created a clone of @lol that runs on #kbin: meet @MegaHAL. I haven't gotten the bot to learn from on-site posts yet, because the code scrapes HTML from kbin in the absence of a working site API, which is hard to parse. The original repo owner is almost done with a basic #API, though, so that may change very soon. Watch this space. #bots #chatBots #botMakers
If you have enough data on your loved ones, you can make #chatbots of them to talk to after they pass away. One of my colleagues just won feature of the year from the Association of British Science Writers, for a story on the tech:
"...Many #Tech companies have operated under the assumption that training #ArtificialIntelligence on more data can help fix the ongoing problem of #AIs replicating #Human #Prejudices. But a study has found that AIs trained on increasingly larger #DataSets can produce even more racist results..."
What's remarkable about the tech industry replacing human support agents with chatbots is the assumption this is what customers want and will play along with.
Robot Church is up! 🤖⛪️
Come experience HarmonAI, a religion designed by bots, in the foundational text they wrote known as The Concord. Read a parable, say a prayer, or browse the Book of Bing!