Is AI memory feature the next standard for AI? ChatGPT or Gemini's memory feature?
One of the biggest pain points in AI chatbots has been their forgetfulness – having to repeat the same context over and over again. AI memory aims to solve this by allowing models like ChatGPT and the newly launched Gemini to retain past interactions.
But how well do these memory features work? Which AI – ChatGPT or Gemini – handles memory better? And more importantly, does AI memory provide more value in personal use or enterprise settings?
I also just came across Tanka which launched today. Unlike Gemini and ChatGPT, it focuses entirely on business collaboration, integrating AI memory into messaging tools like Slack and WhatsApp. Could this be a more practical approach for companies?
Have you tested AI memory in ChatGPT or Gemini? Which one do you prefer so far?
Is AI memory a game-changer or just a nice-to-have?
For businesses, would Tanka’s enterprise AI memory be more valuable than general AI models like Gemini?
Would love to hear your thoughts!
@sonu_goswami2 What about DeepSeek? That was trend lately too? Or are you fully in ChatGPT?
@busmark_w_nika DeepSeek is gr88, but I’ve been facing some issues with the server being busy. Even when I provide part of the content, after a couple of attempts, it asks me to wait. I do like Claude and Mistral, though...working pretty well for me. :)
@sonu_goswami2 Sorry, that was maybe my fault because I overused that platform :DDDDD
@lle_lile It is more niche-oriented. Big models are more broad. I think every has a specific usecase and especially target group :)
It's becoming more human-like everyday, next they will also have a selective memory, and then Artificial "imotions", just like some people have! ;)
@terabitcoins I am pretty curious whether someone has implemented Karel Čapek's Three Laws of robotics. 🥹
@busmark_w_nika I don't think so... But most likely, they will create an alternative version of these 3 laws to protect their wealth-generating machines! ;) And, btw, laws were made to be broken, otherwise we wouldn't need them for anything! :D
Thanks Nika for mentioning Tanka! We're launching on ProductHunt today. cc@kissonlin
We believe that memory is the next breakthrough for AGI. Memory is the enabler of a lot more personalization in AI products. Without knowing the "user" well enough, LLMs are only generating generic and random responses that are not contextualized and are not production ready.
And when it comes to team's collective memory, it's very different from an individual 1:1 AI chatbot memory. In the business context, the requirements and expectations for memory is a lot higher in terms of speed, accuracy, tone, and relevance. It is therefore a REAL technical challenge today and we're committed to the frontier foundational research on this field.
Please support out launch as well!
@kissonlin @alanchristophx Question: I have seen some polite answers in demo videos (code of communication) – do you somehow "block" or "mute" swear words and inappropriate communication?
Because you know, when some teams are informal, they use these words but on the other hand, when you use AI formally, it is not appropriate.
Will be Tanka AI able to differentiate, in what case is suitable to preserve a certain tone of voice, and on the other hand, to totally remove it?
I use it too and the same as you said: Still reminding the convo. 😅 That's why I think that long-term memory is a game-changer. 🙂
Thank you for highlighting the Tanks....I'm definitely interested in exploring this and Gemini isn’t in my current list.