In today’s news:
- On This Day
- 2 AI Truths and a Lie
- Meta announces CM3Leon
- Stability AI releases Stable Doodle
- Google’s Bard chatbot finally launches in the EU
- More NextTech stories
- Graph of the Week
- NextTech Mergers, Funding, and Acquisitions
- The Latest NextTech reads from LXA
This week's NextTech news is trying hard to stay positive.
We've got a good bit of doom and gloom to start your Friday.
Modern Day Luddite and AI theorist Eliezer Yudkowsky believes AI has the potential to change the world. But not in a good way.
"I think we're not ready, I think we don't know what we're doing, and I think we're all going to die," Yudkowsky said on an episode of the Bloomberg series "AI IRL."
He's got a specific suspicion of OpenAI's GPT-4. "The state of affairs is that we approximately have no idea what's going on in GPT-4," Yudkowsky claimed.
Eliezer joins some big names in their reluctance to welcome AI easily. In fact, non-other than OpenAI CEO and survivalist prepper Sam Altman fears his own creation. Dr Frankenstein Sam Altman has reportedly lost sleep over the release of ChatGPT and the potential of 'evil' algorithms.
So, join the club Eliezer.
🤫 One AI Project and Two Lies
Play along at home. Out of these three wacky AI projects, two are fake, and one is real. So, which of these new AI projects is unreal, and which is ugh, for real?! (Check at the bottom of the newsletter for the reveal!)
Hair We Go
On the Fringe
Colour Me Surprised
📰 Meta Announces CM3Leon
Meta, the parent company of Instagram and Facebook, has unveiled a new AI tool called CM3leon, which the company is billing as a “state-of-the-art generative model for text and images.”
Why they've used Leet speak in 2023, but who knows what's going on in the mind of Meta? But clearly, it's being run by 15-year-olds. In 1982.
CM3leon is a multimodal model, which means that it can generate both text and images. It is trained on a massive dataset of text and images, and it can be used to create a variety of different outputs, including:
- Text-to-image generation: CM3leon can be used to generate images from text descriptions. For example, you could give CM3leon a text description of a cat, and it would generate an image of a cat.
- Image-to-text generation: CM3leon can be used to generate text from images. For example, you could give CM3leon an image of a cat, and it would generate a text description of the cat.
- Text-to-text generation: CM3leon can be used to generate text from other text. For example, you could give CM3leon a text description of a cat, and it would generate a different text description of the cat.
CM3leon is still under development, but Meta has said that it plans to release the model to the public in the future. The company has also said that it plans to use CM3leon to improve its existing products, such as Instagram and Facebook.
CM3leon is a significant development in the field of artificial intelligence. It is one of the first multimodal models that can generate both text and images with high quality. This could have a number of applications, including creating new forms of art and entertainment, improving the quality of customer service and developing new educational tools.
📰 Stability AI Releases Stable Doodle
Right, we've got a CM3leon competitor in the second part of this newsletter.
Stability AI recently released Stable Doodle, a sketch-to-image tool that uses the company's Stable Diffusion AI model to turn simple drawings into high-quality images. Stable Doodle is available on the Clipdrop by Stability AI website and app (iOS and Google Play).
To use Stable Doodle, users create a basic sketch with a mouse, choose an art style, and click 'generate'. The tool is designed to be intuitive and easy to use, even for those with no prior experience with AI-generated images.
Stable Doodle offers a variety of art styles to choose from, ranging from realistic to creative. Users can also customize the output image by adjusting the brightness, contrast, and saturation.
The quality of the output images depends on the detail of the initial drawing and the descriptiveness of the prompt. However, even simple sketches can be turned into surprisingly realistic images.
Stability AI envisions Stable Doodle being used by a variety of people, including designers, illustrators, and hobbyists. The tool can be used to create everything from concept art to marketing materials to personal artwork.
Stable Doodle is still under development, but it has the potential to be a powerful tool for anyone who wants to create high-quality images from their sketches.
Here are some of the key features of Stable Doodle:
- Uses Stable Diffusion AI model to generate high-quality images from sketches
- Offers a variety of art styles to choose from
- Allows users to customize the output image
- Easy to use, even for those with no prior experience with AI-generated images
📰 Google’s Bard Chatbot Finally Launches in the EU
Google Bard, the large language model (LLM) chatbot developed by Google, has recently launched in Europe and Brazil, marking its expansion beyond previous territories. Originally delayed in the EU due to data privacy concerns, the LLM is now available to users in these regions, offering its array of capabilities.
If you're here, you probably already know what Bard is. But let me give you a summary of the tech, it's my job after all.
This versatile chatbot has been designed to perform various tasks, such as answering questions in an informative manner, generating creative text formats like poems, code, scripts, musical pieces, emails, and letters, as well as providing language translation and crafting diverse creative content.
By utilising an extensive dataset of text and code, Google Bard can produce human-like responses to a wide range of queries.
The introduction of Google Bard in Europe is a notable development in the AI field, being the first LLM accessible to EU users. Its potential applications include improving customer service, creating innovative educational tools, and facilitating smoother communication between individuals.
Available in over 40 languages, including English, French, German, Spanish, Italian, Portuguese, and Dutch, Google Bard aims to cater to a broad audience. Users can conveniently access the chatbot through the Google Search website or the Google Assistant app.
As Google Bard remains in the developmental phase, the company is continuously working on refining its accuracy and performance. The neutral stance on the launch reflects the ongoing evolution of this advanced AI tool and its potential impact on various aspects of our digital interactions.
📰 OpenAI Vice President Denies They're Making ChatGPT Stupid
Seems like ChatGPT is going through its teenage phase. Being smart is for nerds, after all.
But OpenAI denies ChatGPT is losing brain cells - they're keeping their kid in school.
Ilya Sutskever, the vice president of OpenAI, refuted allegations claiming that ChatGPT, the company's large language model (LLM) chatbot, has become "stupider."
He clarified that no modifications have been made to the model to reduce its intelligence. Instead, any perceived decline in performance may be attributed to increased user activity, which can reveal previously unnoticed issues.
Dall.e Prompt: a robot at school, being handed a test paper with an F grade, oil painting
Users of ChatGPT had reported lower-quality responses in recent months, prompting concerns.
However, Sutskever assured that OpenAI consistently monitors the model's performance and makes changes solely to enhance its capabilities, not to diminish its intelligence.
It is crucial to recognise that ChatGPT remains in the developmental stage, making it prone to errors. Nevertheless, Sutskever emphasised the company's dedication to refining ChatGPT to be both highly intelligent and reliable.
OpenAI's latest LLM, GPT-4, represents their most advanced development yet. While praised for its ability to generate realistic and informative text, GPT-4 has faced criticism for occasionally producing hallucinated information and citations.
Sutskever highlighted that OpenAI is continuously working to improve GPT-4's performance and that all changes aim to enhance its intelligence, not diminish it.
📰Actors Say Hollywood Studios Want Their AI Replicas
Yep, this is a real story. No, you're not living in an episode of Black Mirror. Kind of.
Leaders of the striking actors' union, SAG-AFTRA, criticised the Hollywood studios on July 13, 2023, for offering an AI proposal that would allow studios to replace background actors with their digital replicas in perpetuity in return for a single day's pay.
The proposal, which was reportedly made by the Alliance of Motion Picture and Television Producers (AMPTP), would allow studios to scan the likeness of background actors and then use that scan to create an AI-generated replica of the actor.
Dall.e Prompt: a crowd of background actors in period clothing, with one robot, oil painting
The studios would then own the replica and could use it in any future project, without any further compensation to the actor.
SAG-AFTRA President Fran Drescher called the proposal "an outrageous attack on the livelihoods of background actors" and said that it would "gut the background acting community."
She also warned that the proposal could set a dangerous precedent for other types of actors, including principal actors and stunt performers.
The AMPTP has not yet responded to the criticism from SAG-AFTRA.
However, the proposal has raised concerns among many actors and other creatives about the potential for AI to replace human performers in the entertainment industry.
Here are some of the potential implications of the proposal:
- It could lead to the loss of jobs for background actors.
- It could set a precedent for other types of actors, including principal actors and stunt performers.
- It could have a negative impact on the diversity of the entertainment industry.
- It could lead to the creation of a new class of "digital performers" who are owned by the studios.
Spain's antitrust authority has fined Amazon and Apple $218M for unfairly restricting competition around the reselling and marketing of Apple kit on the eCommerce marketplace.
SoftBank led a $65 million investment into Tractable for car and property damage appraisals using AI
In the last few years, over $33B has been invested into A.I brain implants.
💰Graph Of The Week
Even in its early days, Threads seems to have lured some engagement from Twitter, with a 5% decrease in Twitter's web traffic during Threads' initial peak activity period, per Similarweb. That reduction in traffic could cost Twitter nearly $75 million in worldwide advertising revenues this year and $140 million in 2024.
✍️ NextTech Mergers, Funding, and Acquisitions
Who's making dough, who's laying low, and who's in a constant state of "Oh, God, no"? It's time to find out, with LXA's NextTech News Round.
💰Causaly, an AI platform for discovering and testing drugs, raised $60 million.
💰Preply, the language app known for its live tutors, closes out Series C at $120M
💰Futureverse, an AI + metaverse platform, raised $54 million.
💰FedML raises $11.5M
💰Cognaize, a platform for financial AI applications, raised $18 million.
💰Neura Robotics picks up $55M to ramp up in cognitive robotics
💰Nomic, which helps devs deploy leading AI models locally, raised $17 million.
⚙️Meta releases Llama 2, a more ‘helpful’ set of text-generating models
💰Figure, the builder of humanoid robots, raised another $9 million from Intel Capital.
🤫 One AI Project and Two Lie Reveal:
Let's reveal this with a bang: the "On the Fringe" AI is real!
✒️ The Latest NextTech Reads from LXA
April Fools! You thought this was an article about the top marketing campaigns! Instead, we're now going to go, step-by-step, through my University short film script. It's a rambunctious romp,...Read more