AI & U III

About two years ago I wrote a blog post about AI, basically what it is, what it can do, who has it and where it might be headed. Last year I wrote a second one, AI & U II. Here is a slice of that post, to get us up to date.

*********************************

Today’s update adds a lot to the landscape of this new technology. Turns out that both Microsoft and Google are all over it, and it appears it will be the latest search engine very soon. From this latest article we can soon access this new marvel with Microsoft’s Bing through its browser, Edge. Microsoft tells us their version will be better, more powerful than the current Open AI version. Eh… maybe. I smell some hype in that little slice of ‘info.’ I find it interesting that Microsoft says, “The AI powered engine is still in preview mode, currently allowed to only a few users and reporters. Users can join a wait list to use the engine as it continues to develop and roll out in the next few weeks.”

****************************

How many companies are using it today? Nearly all of them. How many corporations have developed their own version of AI? Here’s a list of the top 10, as of July 2024 –

Apple – market cap of $238.12 (whatever market cap is… dammit Henry, I’m a musician, not a market analyst or a suit!)

Microsoft – $444.75

Google – $197.96

Nvidia – $122.08

META – $677.52

Tesla – $390.18

Adobe – $442.52

IBM – $226.06

PLTR – $80.24

Intel – $19.77

In the past year or so, AI has taken a hard turn toward business functions, chasing the money of course. And the “free art-producing apps” are mostly long gone. I say “mostly” because Dalle-3, Deep Dream AI and Starry AI are still allowing free renderings. And we can now get pictures and designs right from chatGPT! 

Here’s one I just did with chatGPT. I told it “Render a picture of an old man in a cap standing over his name on the Walk of Fame in front of Grauman’s Chinese Theatre. His name is Steve Hulse.” I tried the same prompt on the other two as well. They weren’t even close to chatGPT’s. Barely took 30 seconds and I’m still amazed at the result!  I’ve even got a cap like that!

I’m a big fan of chatGPT. It’s easy, fun and always amazing! If you haven’t yet gotten your feet wet with AI, I highly recommend chatGPT. Just Google it or mention it by name to your local grocer. So far it’s free, no strings. 

I read, as much as I’m able to understand, about the changes AI is going through. The federal government has, along with everything else on its plate, dragged its feet on placing any kind of guard rails on it so far. At the recent AI Action Summit in Paris, 60 countries signed on to a document concerned with the future of AI, wanting it to be safe and accountable for all. J.C. Vance, who represented the U.S. at the conference, refused to sign it. Even China, who is heavily into the advanced development of AI, signed it.

If one is an honest, integrity-driven and empathetic individual, AI is an incredible tool. For the liar, the thief, the power-hungry, the politician, (and no, that was not accidental…) it’s a dangerous tool, dangerous to all of us.

Besides the new direction toward big business, AI is also more available than ever to the masses. The new iPhones, for instance, come with Apple’s version of AI, called Apple Intelligence. Other phones and other new computers also come with their own version of AI. And to give you a short list, I did the obvious… actually asked OpenAI’s chatGPT for the answer. Here it is –

Please list for me the 5 or 6 main versions of AI today, like Bard, etc

ChatGPT said:
Sure! Here are some of the main AI models today:
1 ChatGPT (by OpenAI) – A conversational AI model with multiple versions, including GPT-3.5 and GPT-4. It powers applications like ChatGPT and other AI integrations. ( GPT -5 is a recent addition.)

2 Google Gemini (formerly Bard) – Google’s AI model, rebranded from Bard to Gemini, designed for general AI tasks, search enhancements, and multimodal reasoning.

3 Claude (by Anthropic) – A chatbot focused on safety and alignment, with versions like Claude 1, 2, and the latest, Claude 3.

4 Llama (by Meta) – Meta’s family of open-source AI models, including Llama 2 and the upcoming Llama 3, used for research and development.

5 Mistral (by Mistral AI) – A leading open-weight AI model with powerful generative capabilities, competing in the open-source AI space.

6 Command R (by Cohere) – A model optimized for retrieval-augmented generation (RAG), designed to work well with documents and enterprise applications.

These are some of the most well-known AI models, but many others exist, including industry-specific models and multimodal AI like DeepMind’s Alpha models.

****************************

See that DeepSeek AI? In a recent development, China has introduced DeepSeek, and it has rocked the tech industry, mostly because it does almost everything the big boys’ AI programs are doing, at a fraction of the cost. And its cost of production is what the big deal is about. From what I read yesterday, some of the other big AI’s are scrambling to figure it out, to perhaps re-tool. I suggest they do what I’ve had to do now, countless times – reboot.

It didn’t mention Apple’s version, Apple Intelligence, but who knows? I’m fairly sure they all work way better than most of us will ever need. I’m currently interested in a few of the latest versions of AI on cell phones, which feature a very knowledgeable and friendly voice that you can talk to, talk with. I’ve heard it, experienced it, and it’s scary human. I even saw somewhere that it can become one’s virtual friend. Once again I remind you that I know very little about these new apps, but I can see at a glance they might be very useful… and very dangerous.

So many people these days need a friend of any kind, even virtual. I shudder to think of where that might lead some of us. After all, a sympathetic voice, listening and responding to anything that might be on ones’ mind… well, you get the picture. Sooner or later some of those folk who have been communicating with their virtual friend, that surprisingly real friend who is much smarter than any of the rest of us, are going to cross our path, and I see that, largely, as being much like driving down the highway at 70 mph next to a driverless car. Maybe nothing will happen…

And there, my friends, is my positive and uplifting update on AI today. Our entire government is asleep at the wheel while the programmers are designing new and different functions for AI as we speak. I am but a simple man, but O. M. G. Please, somebody send me back to the ’60’s!

Steve Hulse