Justin Grammens
Welcome readers to another issue of the Applied AI Weekly Newsletter! If you haven't already, I'd like to invite you to follow us on LinkedIn and Twitter
This week we've got articles on everything from AI and the automation of work (our lead story from Benedict Evans, which I highly suggest you read) to the last article, a technical overview for our AI Engineers, on the topic of how to fine-tune your own Llama 2 models!
So, what else is going on in my world? I've just returned from speaking on ChatGPT and Large Language Models at the 2023 Wisconsin THAT conference. It was another awesome event and I have to hand it to the team. If you can attend their next conference in Austin, TX in 2024 I would highly recommend it!
Additionally, here's the run down the next Applied AI Meetup and the podcast episode that has just been released
I hope you enjoy this issue and thank you for continuing to be a subscriber of Applied AI Weekly. Your support means the world to me. Until next week!
News
AI and the Automation of Work
I mentioned it on social media, but this is one of the best articles I've read in a long time. Great perspective on how AI is different from other technological advancements. ChatGPT and generative AI will change how we work, but how different is this from all the other waves of automation of the last 200 years? What does it mean for employment? Disruption? Coal consumption? Read this article to find out!
Social Media Is Fawning Over 19-Year-Old Influencer Milla Sofia - But She's Not Even Real
Milla Sofia looks like your average influencer. She’s 19 years old, blond, and has almost 100,000 followers on TikTok. The twist is that she doesn’t actually exist.
Sofia may claim to be from Finland and post bikini pictures from trips to Greece and Bora Bora, but she’s actually a “virtual influencer and fashion model” generated by artificial intelligence.
OpenAI Can't Identify AI-Generated Text. That's Bad for the Internet.
OpenAI, the creator of ChatGPT and GPT-4, realized this a while ago. In January, it unveiled a "classifier to distinguish between text written by a human and text written by AIs from a variety of providers."
Less than seven months later, the project was scrapped. Let's discuss why!
Sponsor
Software Development Done Right™
Launching a Software Product Is Difficult - We’re Here for You. We help business leaders successfully build and launch their software products with a proven system and a top tier crew of software development experts. Contact us today for your next software development project!
Business
How Will Artificial Intelligence Change Journalism?
Like many discussions about AI, these approaches offer speculative solutions to speculative problems. If not exactly nice to think about, they’re still less grim than the situation immediately in front of newsrooms in 2023. Also absent is the public — the readership, the people who will be theoretically served or pleased or compelled to action by journalism in the coming years.
Is Artificial Intelligence the Future of Wildfire Prevention?
In September 2022, Pano AI announced its collaboration with T-Mobile and Portland General Electric to deploy 5G-connected cameras that were AI-powered and equipped with active wildfire detection in Portland, Oregon. In July 2023, the AI company announced the expansion of its mountaintop camera network with Portland General Electric to six additional locations, bringing the total to 33 across the state.
Healthcare
Implementation: Where AI Programs Go to Die
This article examines the gaps and challenges with implementing and scaling AI/ML models in the healthcare industry and provides recommendations and tips to create lasting impact.
Development
ML Blog - Fine-Tune Your Own Llama 2 Model in a Colab Notebook
With the release of LLaMA v1, we saw a Cambrian explosion of fine-tuned models, including Alpaca, Vicuna, WizardLM, among others. This trend encouraged different businesses to launch their own base models with licenses suitable for commercial use, such as OpenLLaMA, Falcon, XGen, etc. The release of Llama 2 now combines the best elements from both sides: it offers a highly efficient base model along with a more permissive license.