Sponsorship | Go Pro | Accreditations
TOGETHER WITH SUPERANNOTATE X DATABRICKS
Friday’s top story: OpenAI has launched its newest AI model: GPT-4o mini, its smallest, fastest, and most affordable model to date.
🔥 OpenAI reveals breakthrough model
💻 How to fine-tune LLMs with SuperAnnotate x Databricks
🤩 Google AI to transform Olympics
🏢 How to automate your office with AI
📖 How to create a company policy handbook with ChatGPT
🛑 Meta boycotts the EU!
Read Time: 5 minutes
🫤 AI stocks were mixed on Thursday with NVIDIA outperforming the pack, finishing the day green, positive 2.6%. As stated yesterday this area was where buyers stepped in to defend the price at the end of June and yesterday proved to be a similar story. However, Friday’s price will be crucial to determine if the bullish overall trend can continue or if a sustained reversal is possible. Learn more.
Our Report: OpenAI has launched GPT-4o mini, its smallest AI model, which costs less than its full-sized models, is more capable than GPT-3.5, and is faster, more cost-efficient, and smarter than industry-leading small AI models.
🔑Key Points:
GPT-4o mini scored 82% in the MMLU (Measuring Massive Multitask Language Understanding) benchmark test, outperforming mini AI models Gemini 1.5 Flash (scored 79%) and Claude 3 Haiku (scored 75%).
During pre-launch testing, tech start-up, Ramp, used it to extract data from receipts and found that it cost 60% less to run than GPT-3.5 Turbo, and was 2X faster than GPT-4o and GPT-3.5 Turbo.
It’s available now for developers and ChatGPT consumers and will replace GPT-3.5 for general use (although developers can still access it, via an API, but OpenAI has warned that it will sunset it, over the coming weeks).
🤔Why you should care: OpenAI has released this smaller, cheaper, and faster AI model to provide developers with a lightweight, more affordable option for high volume, simple tasks to make “AI more broadly accessible to people,” aligning with its overall mission to advance “artificial general intelligence (AGI) in a manner that benefits all of humanity.”
Together with SuperAnnotate x Databricks
Date: July 30th, 10 am PST / 7 pm CET
While most companies have large amounts of data, they are often unprepared for LLM fine-tuning. During fine-tuning, LLMs learn to mimic the style and content of the data used, and it is therefore important that this data represents the intended model behavior exactly.
During this webinar, we will explore how you can leverage SuperAnnotate with Databricks to refine your data, fine-tune large language models, and thoroughly evaluate the results to make well-informed deployment decisions.
Meco is a distraction-free space for reading and discovering newsletters, separate from the inbox.
ObjectRemover deletes unwanted objects from images with AI
EditApp is for creating and editing images on mobile devices
OpenRead gives AI-powered document analysis
GenPictures converts text to images with AI
Our Report: Google has announced a partnership with Team USA and NBCUniversal (NBCU)—making it the “official AI sponsor for Team USA” and the first tech company to sponsor the team—which will see it integrate its AI-powered features during the coverage of the Paris 2024 Olympics broadcast, starting on July 26th.
🔑Key Points:
The partnership will see NBCU leverage Google Maps to showcase 3D views of venues like the Versailles, Stade Roland Garros, and the Aquatics Centre, and include details about what events will occur at these locations.
Commentators will use Google AI Overviews to answer Olympic-related questions, and US athletes will use Gemini, Google Lens, Circle to Search, and Google Maps Immersive View to explore Paris in social media videos.
The partnership will use Google technology to showcase athletes’ stories, bolster NBCU’s coverage of the event, and enhance the viewers’ experience, showing how Google can help fans explore and connect with the Games.
🤔Why you should care: NBCU has already demonstrated its commitment to using AI to enhance its coverage of the Olympic Games with plans to use an AI-generated version of the voice of Alan Michaels—the iconic American sports commentator—to produce personalized Olympic recaps.
They say AI will simplify everything – but the endless lists of 1,001 ChatGPT ideas make things more complicated, not less.
That’s why we created The No-Cost, No-Code AI Toolkit, a carefully curated guide to the most powerful tools and best use cases for AI in your office workflow.
We’ve combed through all the options, cut out the distractions, and applied these automations to our own $1MM/year tech business – and now they’re ready, for nothing, with zero coding knowledge required.
Type this prompt into ChatGPT:
Results: After typing this prompt, you will get a handbook outlining your company policies and procedures to guide your employees.
P.S. Use the Prompt Engineer GPT by AI Tool report to 10x your prompts.
Meta will not be launching its advanced AI model, Llama 3—which has multimodal capabilities, meaning it can handle video, audio, images, and text—to the European market because of “unpredictable” EU regulators.
EU customers can use the text version of Llama 3, but this decision stops companies, planning to use the multimodal version in their products, from offering them to one of the biggest economic markets.
Meta’s decision follows a similar one made by Apple, who said it probably wouldn’t launch its AI platform–Apple Intelligence–to the EU due to concerns surrounding the Digital Markets Act.
Watch this interview with OpenAI CEO, Sam Altman and Airbnb co-founder, Brian Chesky on deep fakes, ChatGPT, and how AI is evolving.
Hit reply and tell us what you want more of!
Until next time, Martin & Liam.
P.S. Don’t forget, you can unsubscribe if you don’t want us to land in your inbox anymore.