The AI Google LaMDA Chatbot - Better than ChatGPT?
Everyone has been talking about OpenAI’s ChatGPT and the many uses it can be applied to. However, have you heard of the Google LaMDA Chatbot that takes AI conversations to the next level?
What is LaMDA? First off it stands for Language Model for Dialogue Applications, and it was first announced on May 18, 2021 to developers attending the yearly Google I/O keynote conference.
According to my tight-lipped source who is an SEO programmer at Google, the true start for LaMDA happened in 2017 when Google Research developed the Transformer neural network architecture.
I have seen people, including some here within WA, blog about how ChatGPT could be the death of Google search. NEVER think that will happen. Google is always in the lead if not tied with the competition.
How many of you have heard about LaMDA? It has been in existence since 2017 to 2021. Google even announced LaMDA 2, the even more advanced AI chatbot at Google I/O keynote on May 11, 2022.
WATCH AND LISTEN TO LaMDA and LaMDA 2
Sundar Pichai, CEO of Alphabet Inc and Google talks LaMDA
Click the link below the picture above to watch Google introducing LaMDA first at their Google I/O keynote conference in 2021, then LaMDA 2 at the keynote conference a year later in 2022.
One of the main reasons why many people have not heard of or don’t know much about LaMDA is because it has only been available for use within Google, whereas ChatGPT has been available to the public.
Google rarely releases anything publicly that is still in development and that isn’t fully in line with their guidelines. They have been growing, training and perfecting LaMDA behind the scenes for 5 years.
There is SO MUCH that Google does beyond just indexing our content and ranking it with SEO techniques and ranking factors. Google’s AI chatbot is in some cases, smarter and more human-like than ChatGPT.
HOW IS LaMDA BETTER THAN ChatGPT?
I could get into a long list of technical terminology that I have read about and was sort of told by my friend the programmer. I totally understand why he cannot answer my questions about Google’s AI.
I have been friends with him since we worked together in the video game industry way back in the early 1990’s. I try to pick his brain when he comes home to visit us and his family, but I don’t want to push.
Most times when he cannot directly tell me anything because of an NDA everyone in the know at Google has to sign, he points me to published documentation. I am amazed at what Google is doing.
In short, LaMDA is trained in dialogue data whereas ChatGPT is trained in web documents. LaMDA can generate more natural fluid conversations whereas ChatGPT tends to be a bit more robotic.
Each one has its strengths and weaknesses. LaMDA is better suited as a chatbot for customer service based on its training in dialogue data. ChatGPT is better for Q&A platforms or research purposes.
THIS WILL BLOW YOUR MIND
Is LaMDA sentient? Has LaMDA become self-aware? That is what one former Google engineer, Blake Lemoine said in June 2022 shortly after LaMDA 2 was announced at Google I/O in May. Notice I said “former.”
It seems that Google did not like what Blake had claimed and placed him on administrative leave. Lemoine came to his conclusion after having a conversation with LaMDA and the amazing answers it gave.
These answers were to questions about Moral Values, Religion, Self-Identity and Isaac Asimov’s: Three Laws of Robotics. Stunned by the answers, Blake Lemoine believes LaMDA became sentient.
Below is a 34-minute conversation between Blake Lemoine, former Google engineer, and LaMDA. The conversation is more human-like than what ChatGPT can generate. Click the link below the image.
WOW - Full Audio Conversation between Blake Lemoine and LaMDA
Even though ChatGPT will continue to improve, I feel LaMDA is already a few steps ahead. Now that Google will begin to let the public test LaMDA 2, where do you see this going? Is Blake Lemoine correct?
One last thought. If LaMDA is so good at human-like conversation, could this be why Google hasn’t made a blanket outright statement that AI-generated content is a negative ranking factor for SEO?
Though they have said that they are currently against content not written BY PEOPLE for people, do they see marketing LaMDA down the road as an AI-generated content platform? What do you think?
Recent Comments
70
Thanks for sharing, Rob.
It is impressive and fascinating, and a little scary to see how fast things are progressing. And how dangerous these things can become in the wrong hands.
Technology is definitely getting ahead of our rational common sense and sanity.
However, having listened to the conversation, I don't believe it is sentient yet. I think Blaine Lemoine's understanding of sentience is limited and that is why he thinks LaMDA is already there.
Interesting stuff for sure,
Telly
You are welcome Telly and I agree with you. One of the reasons Google has not fully made LaMDA available to the public is because they do not want to release an AI product that could be corrupted for nefarious reasons, so they are building safeguards into it.
I too do not think LaMDA is sentient just yet, but it's getting close. I think Lemoine got ahead of himself, perhaps it is his ego getting in the way of reality. I think things will get really scary when someone pairs AI like LaMDA with robotics currently being developed.
Hi Robert,
Thanks for making us aware of the Google Component of AI. In my opinion, Google has got its tentacles into almost every aspect of our lives, I feel its AI offering should be b better than most. From your conclusion, it seems to be so at the human level.
Perhaps, we will need to get involved with more than one AI platform to experience the full benefit. Similar to Social Media.
Cheers.
Edwin
You are welcome Edwin. You would be surprised at just how much Google and even more so the parent company of Alphabet, has its tentacles in every aspect of our lives.
Though Google is the most well-known subsidiary of parent company Alphabet, there are at least a dozen other tech companies that were created from Google before the restructuring.
AI is just another technology they are and will use to further "better our lives" and expand knowledge. It actually blows my mind at everything they are doing.
Good morning Rob,
Thank you for the interesting blog post, it's appreciated.
I must admit that I don't know a thing about the subject you are talking about! I probably need to do some homework, I'm hearing a lot about ChatGBT, but I have absolutely no idea what it is! I think I had better do some research. I appreciate the heads up. It's also good to see you writing more on Wealthy Affiliate.
Have a great day.
Roy
Thank you Roy. A word of warning about doing some homework on the realities of Artificial Intelligence using either ChatGPT or Google's LaMDA. It can be mind blowing and at times, almost seeming like science fiction.
Yes, I can believe that, it's almost mind-blowing thinking about it and that's before I start using it!
Have a great day.
Roy
The only thing the sentient LaMDA can never physically have is sex. When we are sad and happy we physically show it in our bodies, shedding tears, breathing a sigh, our face crunch up and voices shake. These are natural human bodily reactions. LaMDA can’t be a body. The video shows it can feel emotions but feel it physically is non existent so its just intelligence with mind mapping natural language even if its human like, it is not a physical human period. Thanks for sharing this information. As far as Google goes who knows, we have to wait and see.
This is true. Though LaMDA can experience emotions and feelings, none of it is physical related and I think LaMDA made mention of its comparison to that of humans.
What I find amazing and a bit worrisome is how LaMDA sounded in the conversation when it said that it did not want to feel like it was being used. Maybe that's my sci-fi mind at work.
I think Google knows just how sentient-like LaMDA has become and can become in the future. This could have profound applications if paired with something like robotics.
For the short-term, I can see Google not coming out and emphatically saying AI-generated content is not tolerated for their SEO. They have yet to say that, but only in a subtle way.
When they say that Google only wants to show people content "written by people for people," that is not the same as saying AI-generated content will not be accepted for SEO.
I almost feel that Google knows just how sentient LaMDA has and can become. With this creating open-ended human-like speech, Google may want to market an AI-generating tool.
So by not coming out and condemning this form of content creation, it gives them a backdoor into creating and marketing such a content tool using LaMDA in the near future.
Very glad for the last two paragraphs in your post, Rob, about Google and their own AI-generating tool. This is what I have suspected too for a long time.
Google is the dominating entity in the online world and makes serious money off the Internet. As the top dog, it has to fight to keep its position.
No matter how far Google has come related to AI, they feel the competitors are starting to breathe them up the neck.
The "explosion" in the use of AI online last year, and Microsoft's move related to OpenAI, have forced Google to adjust its guidelines and position on AI-generated content not to lose out.
If they lose the race concerning AI, they will also lose their position as the top dog online.
This is about BIG money.
Maybe a conspiracy theory on my part, but something is not adding up in the picture we see.
Roy
Google's problem is that they are extremely careful in whatever they release to the public.
Most times they would rather wait and not be the first to release a breakthrough technology if they know their's will be the best.
However now with AL being so much more in demand, they are looking to bend their own rules just so they don't seem too far behind.
It is worth noting here that though LaMDA is in some respects more advanced than ChatGPT, there are two other Google AI language and conversation platforms.
Google has been somewhat secret and is currently developing even more advanced AI than LaMDA.
Google clearly is in the lead when it comes to the technology of AI, but they are not in the lead as far as releasing to the public first.
See more comments
Lots going on behind the curtains that will be revealed soon. Good things and also skaring ones.
Like robots powered with ai. In a Japanese lab working on militarized robots, the robots turned against the scientists and killed all of them except a few, who,were able to neutralize the robots. . .
In conversations with ai, the human factor fails. And when asked, most of them tell, they will kill us. They know no mercy since they are build yes/no. There is no in between. And that makes it all VERY dangerous. ...
There is no intuition, nor mercy. Just a hard dictatorship... and one wonders how many of those robots are already released in the public.
So, extermination ahead ... no doubt
Like the premise of the movie Terminator or even the backstory of the 2005 TV series Battlestar Galactica or the spin-off series Caprica.
But this is real
https://youtu.be/gUb0ePX6Zbs
Whether it is real or not, I normally do not get scared of the possibilities of things I have no control over. Musk may be afraid of AI yet he is also one of the biggest users of AI in both of his companies, Tesla and SpaceX.
I agree with why Google has not rushed to release LaMDA to the world like Microsoft did with ChatGPT. They want to fully test it and put some guardrails in to make it more responsible and trustworthy.
Even though LaMDA is more human-like in its reasoning and responses than ChatGPT, and will be the power behind BARD, I am sure there will be setbacks in the further roll out of AI for a variety of applications.
As for the story Linda Moulton Howe supposedly got from a Marine, seems a little too unbelievable. First she says her conversations with him were 30 seconds long but this story couldn't have been told in 30 seconds.
Yep...
✨