$6B for Grok-1, Apple + ChatGPT, Satya Nadella Doesn't like the Term "AI"
$6B for Grok-1, Apple + ChatGPT, Satya Nadella Doesn't like the Term "AI"
A few news and apps for you to check out.
Morning y’all!
Back in the office and hitting those keys like it’s my job. Wait, that is actually my job.
When you reduce my work to simply sitting (and standing) and staring at a small screen for hours a day while typing on little black keys it does make you think for a moment about what you’re actually doing with your life.
But, it’s the only thing I really know how to do. ¯\_(ツ)_/¯
Have a great one folks!
※\(^o^)/※
— Summer
Not much in this announcement outside a very large funding round which will enhance Grok-1 and their models within. They are definitely becoming a competitor if only because the financial backing will inevitably produce results. Will they be good ones or really challenge the top players in the space? Let’s watch and see.
Anthropic have unveiled a new method to interpret the inner workings of its LLMs by mapping out features that correspond to a diverse array of concepts. The goal, among many, is safety in AI models.
Apple is getting in bed with OpenAI and still wants Google as an option for the future. Bringing chatbot functionality into the next release of iOS 18 could be an absolute game changer for both with an immediate growth of adoption and use.
Satya doesn’t like the term “artificial intelligence” and the anthropomorphizing of these tools that are, as he says, just tools. Deepak reminds us that for product management and like it can be useful but not an all-in-wonder:
They definitely improve the productivity of PMs, but as for doing a PM job even in simpler areas like defining metrics, they can do a poor job at times.
What is the future of AGI? Ben has thoughts:
However, for every expert that thinks that AGI might now be close, there’s another who doesn’t. There are some who think LLMs might scale all the way to AGI, and others who think, again, that we still need an unknown number of unknown further breakthroughs.
More importantly, they would all agree that they don’t actually know. This is why I used terms like ‘might’ or ‘may’ - our first stop is an appeal to authority (often considered a logical fallacy, for what that’s worth), but the authorities tell us that they don’t know, and don’t agree.
I like the point in the second block the best: No one actually knows. And I think that’s entirely ok because these types of breakthroughs can never be fully predicted.
Apparently LLMs like GPT-4 can perform financial analysis better than humans in terms of future earnings:
We investigate whether an LLM can successfully perform financial statement analysis in a way similar to a professional human analyst. We provide standardized and anonymous financial statements to GPT4 and instruct the model to analyze them to determine the direction of future earnings. Even without any narrative or industry-specific information, the LLM outperforms financial analysts in its ability to predict earnings changes.
That’s cool. Let’s make some money.
Here are 3 things that will make your ChatGPT prompts better:
Reframe your topic into a question
Use abstraction to anchor it to a larger question
Put your topic into 3 lists
Get the good here.
A framework for building custom AI Copilots in-app AI chatbots, in-app AI Agents, & AI-powered Text areas. Get the React components here.
Here’s the “best” tool for creating backgrounds using AI. Don’t take my word for it.
AI presentation maker that you can use for your next meeting or class. It might save you a few minutes that you don’t have and that would be enough.
Build GenAI apps without managing vector DBs and the like.
A virtual dressing room via AI. I see more of these types of apps coming online because they are so useful!
Another AI copilot that’ll add features to your existing SaaS product.
This smart library will help make all of your knowledge searchable.
And that’s it friends! Have a great day!
※\(^o^)/※
— Summer