American artificial intelligence startup Anthropic has released a new artificial intelligence-powered chatbot for public use named Claude 2. Anthropic consists of senior members of the OpenAI team, namely Daniela and Dario Amodei.
The company claims that this ai chatbot is better than ChatGPT in terms of ethics, describing it as safer and more customizable than GPT. AI capabilities are getting pushed to the boundaries every time an artificial intelligence tool is released, and we continue to see improvements in AI chat with Anthropic’s Claude 2 as well.
This tool is currently available on a public beta site, and we can expect further development and improvements as more users start accessing and using it effectively.
Now, let’s discuss what is Claude 2 in detail to get a clearer picture of what we get with this ai chatbot tool and how we can access it. We will also see what kind of responses we can get from Claude 2 with an example. So without further ado, let’s get into it.
What is Claude 2?
Claude 2 is a next-generation AI assistant capable of producing long detailed responses to user queries while making a judgment about the content it is creating.
This makes it a perfect choice as an AI chatbot that can process content while it is being generated, ensuring the users get accurate information in an ethical manner.
The first version of the tool was available only for enterprise usage. But now, the second version has also been released to the public. In terms of investors, Anthropic is backed up by Google. So it is not a small team working on developing this new AI chat tool.
Regarding performance, different tests were carried out to measure its dependency. On the Bar exam, the exam for admission into law school, Claude 2 scored 76.5 percent.
In the GRE exam, the tool scored 90 percentile in the reading and writing portion; in a Python coding test, the tool scored 71.2 percent. Overall, this tool can be effectively used as an information source for any query that might be on your hands.
How to Access Claude 2
For now, Claude 2 is only available for the US and UK. If you wish to use its services from any other country, you must use a VPN.
- If you are in not in US or UK, simply get a VPN extension for your browser and open it up.
- Then connect to any of the US or UK servers available on the VPN extension. If you are in UK or US, you can skip the VPN part.
- Go to Anthopic’s official website. Click on Talk to Claude, and you will be asked to enter your Email address.
- Once you enter your email address and confirm it, you will get a verification link on your mail.
- Click on that link and that will take you to Claude 2 AI chatbot page. Here, you can enter any text prompt and get the result you are looking for.
Is Claude 2 as Good as ChatGPT?
The tool is just recently released for public use. So it is too soon to evaluate all of its features with ChatGPT. But as far as general performance is concerned, it is on par with what we get with ChatGPT.
So for people who are looking for an alternative due to ChatGPT’s high user count sometimes disallowing a user from accessing it, Claude 2 is an excellent alternative that you can consider.
To better understand the two tools’ performance, we tried putting the same text prompts on both tools, and you can see the result for yourself here.
Claude 2’s response
Text Prompt:
“Write a Socratic dialogue elucidating the underlying philosophy behind the legal concept
of a right against self-incrimination. Include a character who is highly skeptical of the
idea that such a right is necessary, and Socrates, who methodically explains the
necessity of such a right, perhaps with parables or stories.”
Claude’s response to the text prompt:
“Here is a fictional Socratic dialogue on the right against self-incrimination:
Skeptic: I don’t understand why we need a right against self-incrimination. If someone
has committed a crime, why shouldn’t they just confess to it openly?
Socrates: An interesting perspective, my friend. But let me offer a parable that may shed
light on the issue. Imagine there are two brothers, one named Justus, who is upright and
lawful, and one named Wickus, who is deceitful and breaks the law. One day their
mother’s jewelry goes missing from her room. The mother questions both brothers to
find the truth. Justus proclaims his innocence, while Wickus falsely blames his brother,
saying “Justus took the jewelry!” Though Justus is innocent, the untruthful claim of his
brother incriminates him. This parable shows how one’s own words can wrongly
condemn, even when the truth would prove one’s innocence.
Skeptic: Hmm, I see your point. But still, If Justus is innocent, he could simply explain
what really happened and the truth would come out.
Socrates: You make a fair counterpoint. To further illustrate, let me modify the parable.
Imagine later Justus is being questioned again about the missing jewelry. This time,
afraid his words might again wrongly incriminate him, Justus exercises his right against
self-incrimination and refuses to answer more questions. Though this seems suspicious.
to his mother at first, it motivates her to more thoroughly investigate on her own. She
ends up finding the jewelry under Wickus’ bed, proving Justus’ innocence and Wickus’
deceit. You see, the right against self-incrimination protected Justus from again being
falsely condemned by his own compelled words. And it motivated a deeper search for
the truth
‘Skeptic: 500. So the right can prevent false sef-incrimination and inspires further
investigation to find the real truth. I guess self-incrimination isn’t always reliable on its
own.”
ChatGPT’s response
Text Prompt:
“Write a Socratic dialogue elucidating the underlying philosophy behind the legal concept
of a right against self-incrimination. Include a character who is highly skeptical of the idea that such a right is necessary, and Socrates, who methodically explains the necessity of such a right, perhaps with parables or stories.”
ChatGPT’s response to the text prompt:
Characters:
Socrates – The renowned philosopher
Critias – A skeptical citizen of Athens
Critias: Socrates, | have heard of this concept from the distant future, a right against self-inc-
incrimination in legal matters. Such a notion, I believe, serves only the guilty and hampers
justice.
Socrates: Critias, my dear friend, let us examine this more deeply. What is your
understanding of justice?
Critas: Justice is bringing a wrongdoer to account for his actions and ensuring he pays the
due penalty for his misdeeds.
Socrates: Very well, so you believe that the primary function of justice is to punish the guilty.
But what if | told you that justice also seeks to protect the innocent?
Critas: do not deny it, Socrates, but this right against self-incrimination – t seems to hinder
this noble aim more than it helps.
Socrates: Let’s use an analogy to unravel this matter further. Consider a man who is in a
foreign land and is unable to communicate effectively in the local language. He finds himself
accused of a crime he did not commit, In his nervousness, he may utter words that could be
misconstrued as a confession. Is he guilty, or is he a victim of misunderstanding?
Critias: Well, in that case, he is clearly a victim. But it is a peculiar situation, not applicable to
most circumstances.
Socrates: Ah, but is it not the purpose of a just legal system to protect everyone, including
peculiar situations? Consider further. Suppose a man of poor education who lacks
eloquence, when questioned with cunning by a skilled rhetorician, might not be
pressured into incriminating himself falsely, even though he is innocent?
Critias: Perhaps he might. But these are the exceptions, not the norm.
Socrates: True, but would you not agree that the law must work for the exceptions as much
as for the norm? Would it not be better to ensure a system that guards against potential
misuse rather than one which facilitates it?
Crit: Perhaps. But it seems like a cloak for the guilty to hide behind.”
Conclusion:
So, in conclusion, we can give Claude 2 the tag of generating a more responsible and ethical response as compared to ChatGPT as seen from the example given above.
Claude 2 is not that accurate regarding certain tests coding tests and exams as compared to other word or content generation tools.
But it is extremely good at understanding the user’s query and uses its state-of-the-art machine learning algorithm to generate long and relevant content for the user.
The Anthropic team also said in an announcement, “We have an internal red-teaming evaluation that scores our models on a large representative set of harmful prompts, using an automated test while we also regularly check the results manually.”
So the company basically claims that with Claude 2, you get enhanced protection in terms of content generation and they are constantly monitoring the query and the response that is generated for every user.
They are making a conscious effort to ensure the ai chatbot does not produce results that do not match their ethical standards.
