Facebook AI Blender the bot that lies

His researchers Facebook AI announced today Blender, an advanced open source chatbot model. The new model is supposed to be "more human" than previous versions and can have better discussions.

Based on the first signs it seems to be as good as the best competing bots from Google and Microsoft.

The Facebook φέρεται να χρησιμοποίησε ένα τεράστιο σύνολο πληροφοριών και εκπαίδευσε το AI. Τα blender chatbots είναι ανοιχτού κώδικα, κάτι που σημαίνει ότι δεν μπορούν να εκπαιδευτούν απλώς για να απαντήσουν σε μια συγκεκριμένη σειρά ερωτήσεων, όπως αυτές που βλέπετε σε ιστότοπους.

These bots are theoretically capable of talking about any topic.


A post from the company blog states:

This is the first time a chatbot has learned to combine several chat skills, such as the ability to have a personality, discuss almost any topic and show empathy - in natural chat streams.

Our new recipe incorporates not only large-scale neural models with up to 9,4 billion parameters - or 3,6 times more than the largest existing system - but equally important skills combination techniques and detailed discussion.

The number of parameters used by Blender plays a big role in how convincing each AI discussion is. AI requires more computing power, but it produces significantly better results.

According the page of the Blender project:

Good conversation requires a set of skills that a skilled interlocutor combines seamlessly: Provide attractive points in the speech and listen to the collaborators, ask and answer showing that he has the necessary knowledge, has empathy and personality, depending on the occasion . It seems that large-scale models can learn these skills when given the right training data and a choice of strategy.

The result is a chatbot that can convincingly reproduce human conversation. In a way, it is capable of passing the Turing Test. This may well represent the most advanced AI chat on the planet.

Η Artificial Intelligence of Facebook δεν καταλαβαίνει λέξη από αυτά που λέει. Δεν “ακούει” κανέναν, δεν μπορεί να καταλάβει τη γλώσσα ή το νόημα πίσω από τις λέξεις. Κάνει μόνο τους συσχετισμούς μεταξύ των δηλώσεων που του έχουν δώσει οι προγραμματιστές του και κρίνουν ότι είναι οι κατάλληλες απαντήσεις σε ένα θέμα.

He may not be "knowledgeable" because he does not actually use a limited database, but he makes correlations that seem to make sense. For example, if you say that your favorite song is one of Smashing Pumpkins, he may say that he likes a different song better. But he has never listened to music. Edits only natural language. He does not understand what sound, images or video are.

He also has no empathy, can not analyze your feelings or express his own. He simply responds with statements that he has been trained to understand as representing the appropriate answers. If, for example, you say you are sad, he will not say "congratulations", or "and time."

Basically, the Artificial Intelligence of Facebook φαίνεται να γίνεται ένας τέλειος ψεύτης γιατί δεν έχει άλλη επιλογή. Η τεχνητή νοημοσύνη δεν αισθάνεται ούτε σκέφτεται, οπότε δεν μπορεί να βιώσει.Έτσι χωρίς εμπειρίες, δεν θα γίνει άνθρωπος. Το bot θα πρέπει να πει ψέματα για να το “παίξει” άνθρωπος.

There may come a time when perhaps humanity regrets its decision to create bots of artificial intelligence. The biggest danger, however, is not the bots who will use their knowledge to harm us, but the people who can do it.

If you want more information about Blender, see the paper from here.

Read them Technology News from all over the world, with the validity of iGuRu.gr

Follow us on Google News iGuRu.gr at Google news