🤔 w/143: ChatGPT Is Given The Power Of Sight & Sound
Plus: Bitcoin's trend to stability, Roblox brings metaverse to Playstation, BigTech's AI features to make life easier and AI misinformation is on the rise.
Hi readers, last week I called out that I’m making some changes to the format of Wiser! It’s a period of transition, so hang in there with me. And also let me know what you think…is this better, worse, you don’t care? Either way, hit the heart button or drop me a comment. It gets lonely with no feedback.
Have a great weekend, Rick
w/Commentary
ChatGPT’s Been Given The Power Of Sight And Sound
OpenAI’s latest updates to ChatGPT are pretty impressive. You can now speak directly to ChatGPT. The AI can hear you, understand you and speak back to you. It’s the experience we were promised with Alexa and Siri that Amazon and Apple never delivered on. This is super convenient, especially if you’re on the go. It’s also quicker than typing. I’ve been rambling into my mobile ChatGPT app all week and it’s a lot of fun, more immediate and remarkably accurate at getting to the heart of what I want to know. Just imagine what this will be capable off when it’s connected real time to the internet. Bye Bye Search!!!
But more impressive than this is that ChatGPT now has the power of sight. Upload a photo from your holiday and ask ChatGPT to describe it, locate it or tell you something interesting about it. Or take a screenshot of your social media analytics and ChatGPT will suggest how you might increase engagement. Or give it a piece of homework and ChatGPT will fill in all the blanks.
Over the week, as well as playing with it myself, I’ve seen a number of examples that illustrate the capability of ChatGPT-4V (the “V” is for Vision and it runs on GPT-4). ChatGPT gave coaching advice based on a a still photo from an American football game. The AI was shown some photos and explained how to improve the lighting, framing and perspective. A hand drawn sketch for a website design was turned into the code to build it. Take a photo of the food in your fridge and ChatGPT will give you recipes for it.
To understand how it does this, I need to introduce to you a new three letter acronym. By now you’re familiar with LLM, which stands for Large Language Models. These AI systems are all about words. Now we have LMM, which stands for Large Multimodal Models. These AI systems are all about images. Just like words, they are trained on gazillions of images, breaking them down into constituent parts and giving each part a label (called a token).
IMHO, image processing is going to be more significant than word processing (a picture paints a thousand words and all that!) It’s fun, exciting, and also scary, all at the same time. And it’s worth remembering that it’s still less than a year ago since we saw ChatGPT for the first time. And it was only six months ago that we got GPT-4!
This pace of innovation and change is unprecedented (which is exciting). But the consequences of this unleashed technology are still unknown (that’s the scary part).
Here’s The Thing: OpenAI has given us multi-model AI for general use. It’s taking us one step closer to a screenless future. If the AI you’re using has the power of sight and sound, logic and language, then you won’t need a screen anymore. The power of computing will become invisible, always there, always on, and always listening.
You can see it already happening. Meta’s Ray-Ban sunglasses have the AI powered sight and sound. Two weeks ago, Naomi Campbell modelled Humane’s AI Pin on the catwalk of a Paris Fashion Show. Humane’s AI Pin is a camera, speaker and listening device without a screen that clips to your lapel. Rewind just announced the Pendant, an all listening AI assistant you wear around your neck that records every single thing you say so that AI can remind you what you had for breakfast this morning.
This time last year, AI wasn’t being talked about much outside of the geeky data science community. Now, it’s everywhere and in everything. Imagine what it’s going to be like 12 months from now! That’s both exciting and scary!
w/Data
Bitcoin Shows Stability in 2023 Despite Declining Enthusiasm in the Crypto Sector
For the most prominent cryptocurrency, the chaos of the 2020-2022 era seems long gone. Bitcoin, which was often criticised for being too volatile, hasn't seen a daily gain or loss of more than 10% this year. That's in stark contrast to 2022 and 2021 when it swung outside this range 9 and 11 times, respectively. All told, the price of Bitcoin has risen more than 60% this year, despite the continued crackdown on major exchanges such as Binance and Coinbase.
Source: Chartr
➜ 87% of US teenagers own an iPhone according to a new survey.
w/News
What Else Is Going On In Tech?
Immersive
Roblox has come to the PlayStation. Why does this matter? Because Roblox, probably more than anyone else, is building the next generation of immersive tech we loosely call “metaverse”. Remember, Roblox is the gaming platform for kids, the next generation of consumers and internet users. Putting it on Playstation brings Roblox to a wider, older audience.
Meanwhile, Microsoft have been given the green light by UK regulators, to complete the biggest deal in gaming history and buy Activision Blizzard for $69 billion.
And the Meta Quest 3 VR headset has hit stores worldwide. With a 40% thinner and more comfortable design, and significantly more computer power than its predecessor, the Meta Quest 3 offers both virtual and mixed reality experiences in a single device. The Quest 3 is receiving positive reviews, which was evident if you watched the Lex Fridman podcast with Mark Zuckerberg. The tech has come on a long way in a year and, for the first time, looks like it has a real chance of wide-scale adoption.
➜ This is interesting: State Of Blockchain Gaming, Quarter 3 2023 illustrates the growing shift from Web2 to Web3.
Deepfake News and Misinformation
“As we go boldly forth into this future, a photo is no longer a visual fact.” - Ren Ng, a computer science professor at UC Berkeley in an interview with the New York Times about AI photo editing tools.
Videos featuring AI-generated voices are flooding TikTok with conspiracy theories and falsehoods, mimicking celebs like Barack Obama and Elon Musk. The threat of fake news and misinformation is very real, and an order of magnitude greater than we saw a decade ago on Facebook and Twitter (remember Cambridge Analytica?). This is proving a challenge for TikTok to detect, take down and label deceptive clips reliably.
Meanwhile, over on Twitter a “surge of disinformation” has overwhelmed users with fake photos, outdated videos, and video game footage misrepresented as actual events in the Israel-Hamas conflict. It’s fair to say that fake news is all over every social media platform, but the thing that makes it worse at Twitter is that Musk has removed most of its content moderation capability. Most notably is a recent removal of a piece of software that automates the authentication of images. The European Union were quick to jump on Twitter, TikTok, and Meta, demanding to know how they’re moderating fake posts about the conflict. Under the Digital Services Act, the EU can fine these firms 6% of global revenues for failing to control misinformation.
FYI: The Washington Post has published a guide to spotting misinformation, focused on tips to help identify when AI has been used to create or edit images, videos, and news stories. My advice, stick to credible sources you’re certain you can trust.
My work on deepfakes: Big Tech Little Tech | Deepfake Movies | Deepfake Ethics
The Utility Of AI
Time spent waiting at a red light is more than just an annoyance, it’s also terrible for the environment. In a 2015 study, scientists estimated that city intersections tend to harbour around 29 times more pollution than the open road. To find a solution, Google created Project Green Light using AI and Google Maps to coordinate traffic lights more efficiently. Two years on and Google say that initial results reduce stops by 30% and cut emissions at intersections by 10%.
Domino’s Pizza is using AI to optimise the ordering of pizza, employee scheduling, and other key operational aspects of its business. Meanwhile, a software and robotics maker called Symbotic is being used by Walmart to provide AI-powered logistics and warehousing as a service.
Researchers at Harvard and the University of Oxford have built an AI system that can forecast new variants of viruses like SARS-CoV-2, HIV, and influenza with the goal of future-proofing vaccines against mutations.
Meanwhile, Researchers in the Netherlands have developed an AI system that can rapidly analyse DNA from brain tumours during surgery. This is important because surgeons have to make real-time decisions about how much of the brain they have to remove during operations that can take many hours. Using the AI, they hope to be able to better identify the healthy areas to leave alone. (It’s more complicated than that, but you get the gist!)
What’s New In Generative AI
Adobe has revealed an experimental tool called Project Fast Fill. It’s a new way to edit videos using AI generation and text prompts that’s a bit like Google's Magic Eraser for photos, but it does the same video! The demo looked impressive and will bring Hollywood-esque video editing capabilities to your desktop! There is a downside…see earlier section about deepfakes and misinformation.
Google has introduced new features to enhance the AI search experience. Users can now search for images and drafts directly within the search engine, making it easier to find visual content and work on unfinished projects. Here’s the thing: Google is prioritising user convenience and efficiency in its approach to maintaining its hold over Search.
YouTube has introduced new AI features to enhance social media marketing and content creation, including A.I. insights for creators, dream screens for photo/video backdrops, and A.I. translations for expanded reach.
On November 1st, Microsoft will introduce Microsoft 365 Chat. This is a powerful new tool that you can train to complete tasks like managing your inbox and planning meetings. Again, it’s an example of convenience and efficiency, using AI to save you real time by taking care of the non-value creating, time-absorbing stuff like summarising emails, drafting replies, auto create to-do lists and schedule your calendar.
Meanwhile, Microsoft owned LinkedIn has introduced new AI features that enhance networking, recruiting, and learning, aiming to make these processes 10 times easier.
w/Productivity
Constant Contact
With the rise of AI, companies are exploring new ways to enhance their customer experience. Constant Contact has been leading the way in the email marketing and CRM industry by incorporating cutting-edge AI technology into its platform.
From AI automations to content generation, Constant Contact offers a range of advanced features to help you create highly targeted, effective campaigns without the hassle of writer's block. Say goodbye to the frustration of content creation and hello to more time doing what you love with who you love. 🙌
I used Constant Contact’s AI content generator to help write this post! So follow my lead and try check out Constant Contact's AI solutions today and see the difference it can make for your business too!
w/Insights
Insights To Make You Wiser!
The Rise of Screenless Computing: A Glimpse into a Future Without Screens
Sam Altman’s WorldCoin: The Convergence Of AI and Cryptocurrency?
FREE TO WISER! READERS: Download your copy of The Utility Of Emerging Technologies: An eBook with 25 case studies of consumer brands using new technologies to engage with customers.
🙏 Show Your Support For Wiser!
Thank you for reading Wiser! If you got value and would like to support what I’m doing, do this:
Forward this email: send it to anyone you know who’s interested in the tech economy.
Make a donation: go to BuyMeACoffee to make a donation in the form of a virtual cup of coffee, they only cost €2 each. And who doesn’t love a coffee, right?
Check out my website: You’ll find links to everything I do at rickhuckstep.com.