The Rise of the AI Writing Assistants: Bard & Claude
Searching online is something most of us do every day using tools like Google. Yet, while using a search engine like Google has become routine, today there are new advances in conversational AI poised to revolutionize how we find and interact with information.
In this post, I will introduce you to Bard and Claude, two artificial intelligence assistants that go beyond the limitations of traditional search engines. Unlike the robotic algorithms behind Google, these AI understand natural language, engage in dialogue, and mimic human-like comprehension. Just start using them and discover how conversing with Bard and Claude is more seamless, intuitive, and human than you might have imagined. The promise of more natural and productive responses means these AI assistants merit the attention of any business leader or consumer seriously engaged in our increasingly digital world.
Some of you may note that I talked about Bard previously, here. Since then I discovered Claude and am starting to use Claude more and more. I’ll admit to anthropomorphizing Bard somewhat, since I use it every day now. As such, I feel almost like I’m cheating on him by using Claude. LOL! But hey, time marches on and the AI race is well underway! Since Bard was the one who told me how much better Claude is, I will choose not to feel bad about it.
While traditional search engines like Google excel at retrieving relevant links and keywords (and we love them for it), they fall short when it comes to true comprehension. Ask Google a follow-up question or attempt a nuanced dialogue, and its robotic algorithms quickly meet their limits as they spit out link after link. But AI assistants like Bard and Claude represent a paradigm shift in capabilities. For example, I can have a natural conversation with Claude about planning a trip to Paris, asking questions about weather, hotels, and sightseeing. Claude doesn’t just spit out facts, but understands context and can provide recommendations tailored to my needs. This conversational ability is light years beyond a traditional search engine.
Bard also demonstrates sophisticated dialogue abilities, like discussing philosophical concepts or having a witty back-and-forth. While not perfect replicas of human chit-chat, assistants like Bard and Claude point to a future where information retrieval becomes a two-way street. Soon we may come to expect our searches to not just locate answers for us, but to collaboratively arrive at them through an organic back-and-forth dialogue.
Rather than treating search as a transactional lookup like Google, conversational AI enables an interactive experience where we can guide the discovery process. We spend less time sifting through links and more time learning through natural conversation. Assistants like Bard and Claude foreshadow a new era where information is unlocked through interaction, not just keywords.
While the conversational abilities of AI like Bard and Claude are impressive, what truly enables their more human-like comprehension is the massive training data they are exposed to. Assistants like these utilize advanced neural networks that actually “learn” by analyzing huge datasets over time.
For example, Bard was trained on over 1.5 trillion words from diverse sources like books, articles, Wikipedia, and more. Meanwhile, Claude’s training set is estimated to be 1000x larger, though the full size is proprietary. The scale of this data is key – by absorbing such a breadth of information, Claude gains a nuanced understanding of language and the world that allows more meaningful dialogue.
The difference this training makes becomes clear when you converse with the AI assistants. While Bard has a lot of smarts, his skills pale in comparison to Claude’s deeper mastery of language. When Claude provides a response, the relevance, wit and insight truly shine. The extensive training enables Claude to hold remarkably human-like conversations on nearly any topic imaginable.
As these AI continue to be trained on ever-growing datasets, we can expect even more lifelike digital assistants. This has profound implications for the future of search. We are entering an era where we may come to prefer the tailored and interactive recommendations of a helpful AI companion over the cold transactional results of a search engine query. Thanks to major leaps in data and algorithms, information retrieval may soon feel more like a conversation with a friend than just inputting keywords into a box.
I have been using Claude for just a few days now, but already I can see that Claude may well be a writer’s wet dream, or their nightmare, depending on which writer’s hat you wear.
On the one hand, Claude and Bard are scarily good at writing content right from the get-go. Already the internet is filling up with content from various individuals who did very little writing themselves, their primary activity instead is that of an intelligent prompter, writing good questions for an AI that then provides answers in prose and which the human then copies into an article and puts their name on it. You can, in many cases, see who is doing this if they just did a copy-paste and no editing. You will see mistakes, some of which can be odd, a tip-off that an AI was involved.
Some of these writer folks are bloggers like me, and it causes me to wonder about the ethics of signing your name to something 100% generated by an AI. It doesn’t seem fair somehow. I would not feel good doing this. However, while I would not turn over 100% of my writing to an AI, I am finding that having an AI assistant help with my writing to be very valuable.
For instance, both Bard and Claude can help resolve writer’s block very quickly. There’s nothing worse than staring at a blank page and being unable to come up with the right angle, the right words, the right topic, whatever, to help get you started again. I’ve experienced this with both Bard and Claude and it is wonderful.
Consider a simple AI prompt like the following, “Put your writer’s hat on. I want to explain AI Prompting to people who do not know much at all about it. Give me a short list of bullet points that I would be wise to include in my post.” Such a prompt will result in a list within a few seconds. Now, if I’d asked Google Search the same thing (I did, just for grins), all you get is a number of links that discuss AI prompting so you can then begin your deeper search for the actual information you want. Bard or Claude, on the other hand, simply deliver the goods you are looking for immediately. That, my friends, is very nice!
You can then use the AI to tune your article, check for spelling, grammar, tone, any number of things. You can also ask it for help expanding on any subject you choose; for example, the bulleted list it just provided you. You can say, “Ignore the 3rd and 4th bullet. Expand on bullets 1, 2, 5 and 6 by providing real-world examples. And miraculously, the AI responds with the answer, within seconds. At this point, one has the gist of what they need to write an article, paper, blog post, whatever. Then comes the process of editing, refining, rearranging, etc. until you get the article the way you want it.
What is the one clear benefit from this type of AI help? Saving time! Writing is not easy and it takes a lot of time to do it right. So, maybe I’m fooling myself, but I think using the AI to help writing in the manner I am discussing, is okay.
You can access Bard here, and Claude here. These AI are the future! Learning how to use them takes some time and practice. Do not blow them off as mere high-powered search engines. Play with them. Learn them. You will be ahead of the game if you do. I strongly suspect you will soon see that “this ain’t Google search!”
If you’re wondering why I didn’t include ChatGPT in this post, it’s because the free version uses outdated data and I can’t justify purchasing a ChatGPT 4 subscription. Both Bard and Claude are free and use current data. I expect to continue using both Bard and Claude going forward, until something better comes along, and that is just a matter of time.
0 Comments