23/6/2024
AI-Related Wizards23/6/2024
AI-Related WizardsStart by listening to our AI-Generated Podcasts
Commentary ➲ Perplexity's Grand Theft AI
Introduction written by Perplexity
Commentary ➲ Is this how blog articles and podcasts will now be created?
Perplexity Pro AI was used to generate this blog article!
Commentary ➲ Caution is Warranted ...
Source Citations and Transparency
How to use Perplexity Pages for Content Generation
Getting Technical ➲ How to convert Perplexity Pages to HTML?
For ease of understanding which parts of this blog article were generated by AI, and which parts were generated by a human (that would be me!) I have applied the quotation style to the text and comments I have added, for an example, see the below text:
"Perplexity has been widely criticised in the news recently (for example: see Perplexity's Grand Theft AI article by The Verge dated 27 June 2024) by both journalists and website owners who have specifically set their robots.txt file to deny permission for Perplexity search bots to scrape their content.
Despite this, compelling evidence has been uncovered showing that Perplexity bots (albeit rogue 3rd party bots) are indeed continuing to scrape content ignoring the specific instructions of the website owner.
I hope and trust that Perplexity can promptly resolve this issue as it represents a major issue.
The path forward necessitates payment for a license to use content, assuming the journalists and website owners are willing to grant one."
Perplexity AI revolutionizes information access and content creation by combining real-time web search capabilities with advanced AI-powered natural language processing.
This innovative platform offers multiple language model options, including GPT-4, Claude 3, and custom Perplexity models, enhancing its versatility for various tasks.
With features like Pro Search for in-depth queries, organizational tools such as Threads and Collections, and a commitment to source transparency, Perplexity AI caters to diverse user needs. The platform's potential extends to transforming legal research, website content generation through Perplexity Pages, and even offers workarounds for HTML conversion. By bridging the gap between traditional search engines and conversational AI, Perplexity provides a tailored, efficient research experience that's reshaping how we interact with and create information.
"The official published version of this blog article is hosted on Perplexity Pages.
The blog article was quick to generate (you read that correctly, I said generate not write ... ) yet the end product is highly polished, user-friendly and incorporates all source links used in its generation.
Perplexity Pro allows you to select one of its proprietary LLMs or from a selection of the latest leading LLMs to experiment & fine tune your results.
This blog article made use of the latest LLM from Anthropic - Claude 3.5 Sonnet as well as Playground and DALL-E 3 for AI image generation.
You are encouraged to take a look at both the simplified version here as well the official version.
Please let us know what you think ...
Is this how blog articles and podcasts will now be created?
We think so ... however, we feel that the most successful content will be the outcome of a hybrid collaborative effort between person and the AI-Generated content.
That is the reason we have added distinctly headed up "Commentary" sections in this blog article, and why we intend to update and republish our Podcasts with real-person perpectives for the same reasons we have added these to the source blof articles.
Click here to scroll down to the bottom of this blog article to listen to our AI-generated podcasts generated from this article as well as one of our related blog articles.
The podcasts use an "AI co-hosted dialogue" format."
Perplexity AI is emerging as a valuable tool for legal research and analysis, offering potential to transform various aspects of the legal profession.
Here's an overview of its impact:
Perplexity AI's ability to process and analyze large volumes of legal information quickly can significantly enhance the efficiency of legal research.
Unlike traditional search engines, it provides coherent narrative responses to legal queries, complete with citations to relevant sources.
This feature is particularly useful for lawyers and law students who need to quickly grasp the key points of complex legal issues.
In case analysis, Perplexity AI has shown promise in accurately summarizing case facts, identifying relevant legal provisions, and extracting key issues.
This capability can help lawyers prepare more effectively for cases and potentially uncover insights that might be missed in manual review.
For contract review, Perplexity AI can rapidly process lengthy documents, highlighting important clauses and potential areas of concern.
This can streamline the contract review process, allowing lawyers to focus on more nuanced aspects of negotiation and drafting.
In legal education, Perplexity AI serves as a supplementary tool for law students. It can help clarify complex legal concepts, provide summaries of important cases, and offer practice in legal reasoning. However, it's important to note that while Perplexity AI can be a valuable resource, it should not be seen as a substitute for thorough legal education or professional advice.
The tool's ability to simplify legal jargon can also improve client communication.
Lawyers can use Perplexity AI to generate clear, concise explanations of legal concepts for their clients, enhancing understanding and client satisfaction
While Perplexity AI offers significant benefits, it's crucial to approach its use in the legal profession with caution.
Legal professionals must verify the accuracy of AI-generated information and use it as a complement to, rather than a replacement for, their expertise and judgment.
As the technology continues to evolve, its impact on the legal profession is likely to grow, potentially reshaping how legal work is conducted and delivered.
"By way of human commentary based on my personal experience in using Perplexity for online legal research, I can confirm that the caution raised by Perplexity itself is indeed warranted.
Even as the accuracy of LLM models continue to improve, as you would expect with any AI automatically generated text (it can never be relied upon as authoritative particularly on legal matters).
The same logic applies for anything legal research findings based on Google search results, even if it contains information presented on law firm websites.
These are all only starting points for legal research!
Any automatically generated summary based on these sources needs to always be treated with caution.
Let's consider the kinds of errors I have personally identified in online legal results generated by Perplexity.
Example 1:
This search answer mentions a court decision called Hall & Hall but did not provide a citation.
When I dug in and investigated the sources used by Perplexity I noted that there were two court decisions called Hall & Hall, once was from 1954 and one was from 1990.
I first looked at the 1954 court decision - the decision made no sense with respect to the legal research and did not consider the Act mentioned in the answer.
As the text mentions the Act (but not the year the Act was passed) I needed to determining when the Act was passed, which turned out to be circa 1986 (from memory).
I was then able to review the 1990 court decision and confirm that it was indeed the correct one and was authority for the legal principle that was being summarized,
At this point, I was able to add the correct citation.
This additional legal research work has only corrected the initial summary, so that it can be relied upon as a starting point to conduct traditional legal research using the trusted proprietary legal research tools traditionally used by lawyers to determine whether or not this statement represents current law.
Example 2:
The same search result cited a court decision from New York in support for a statement that confidently states the decision was from a court in California."
Unlike static language models, Perplexity AI actively scours the internet to provide up-to-date information in response to user queries.
This real-time web search capability allows users to access current information, even for rapidly changing topics or events.
For instance, Perplexity can be used to find the score of a live soccer game, demonstrating its ability to retrieve instant, real-time data.
The platform's daily indexing of the internet further enhances its capacity to deliver fresh, relevant content to users.
Perplexity AI offers robust organizational features to help users manage and structure their research efficiently.
Here's an overview of the key organizational capabilities:
Perplexity AI sets itself apart by offering users access to multiple large language models (LLMs), providing flexibility and diverse capabilities.
Here's an overview of the LLM options available on Perplexity:
This multi-model approach allows Perplexity to leverage the strengths of different LLMs for various tasks. GPT-4, developed by OpenAI, is known for its advanced natural language understanding and generation capabilities. Claude 3, created by Anthropic, excels in reasoning and analysis. Mistral Large, while less well-known, may offer efficient performance for specific types of queries
Perplexity also utilizes its own custom models, which are likely optimized for the platform's unique search and summarization tasks. This combination of third-party and proprietary models enables Perplexity to provide more comprehensive and accurate responses across a wide range of topics and query types.
The platform's ability to switch between these models gives users the flexibility to choose the most appropriate LLM for their specific needs.
For instance, one model might be better suited for creative tasks, while another could excel at technical or analytical queries.
Additionally, Perplexity's multi-model approach extends to image generation, with access to models like Playground, DALL-E 3 and Stable Diffusion XL.
This multimodal capability allows users to generate both text and images within the same platform, enhancing its versatility as a research and content creation tool.
By integrating multiple LLMs, Perplexity can potentially mitigate the limitations of any single model, providing more robust and diverse AI-powered search and generation capabilities.
This approach also allows the platform to quickly adapt to advancements in LLM technology by incorporating new models as they become available.
Perplexity AI's Pro Search feature offers an enhanced, interactive search experience compared to its Quick Search option.
Here's a comparison of the key differences between Pro Search and Quick Search:
Pro Search is designed to provide a more comprehensive and personalized search experience.
It engages users in a conversational manner, asking for additional details to refine its understanding of the query.
This feature allows Pro Search to deliver more accurate and relevant results tailored to the user's specific needs.
Unlike Quick Search, which aims for speed and general information, Pro Search delves deeper into topics, considering user preferences and the nuances of their questions. This can be particularly useful for complex queries or when users need a more thorough exploration of a subject.
The interactive nature of Pro Search helps users avoid the need for multiple searches or sifting through irrelevant information.
By asking clarifying questions, it can fine-tune its responses, potentially saving time in the long run despite the initially longer processing time.
Pro Search's ability to understand context and user intent makes it particularly valuable for research, problem-solving, or when dealing with multifaceted topics.
It aims to provide a complete view of the subject, offering insights that might not be immediately apparent with a standard search.
While Quick Search remains useful for straightforward queries requiring rapid responses, Pro Search positions itself as a more sophisticated tool for users who need depth, accuracy, and personalization in their search results.
Perplexity AI distinguishes itself through its commitment to source transparency and citation.
Unlike many AI chatbots, Perplexity provides numbered references for every line it generates, linking to the original web sources.
This feature allows users to verify information accuracy and explore topics further.
The citations come from various platforms, including academic publications, enhancing the tool's credibility for research purposes.
Additionally, Perplexity's real-time web search capabilities ensure that these citations are current and relevant, making it a valuable resource for up-to-date information retrieval and fact-checking.
Perplexity AI offers powerful features for generating website content through its new Perplexity Pages tool.
This feature allows users to create detailed, well-structured articles on various topics with minimal effort.
Here's an overview of how to use Perplexity AI for website content generation:
Perplexity Pages automatically creates well-formatted articles with citations, making it easy to produce informative content quickly.
Users can further customize the generated content by adding specific subtopics or adjusting the tone to match their website's style.
While this tool streamlines content creation, it's important to review and potentially enhance the AI-generated content to ensure it aligns with your website's unique voice and quality standards.
Perplexity AI currently does not provide a native solution for converting Perplexity Pages directly to HTML for use on external websites. This limitation aligns with Perplexity's business model, which encourages users to share content through their platform rather than facilitating easy export to other sites.
However, there are several workarounds and third-party tools that users can employ to convert Perplexity Pages to HTML for their websites:
Manual Conversion: Users can copy the content from Perplexity Pages and manually convert it to HTML using their preferred text editor or HTML authoring tool.
Markdown to HTML Converters: Since Perplexity Pages are in markdown format, users can utilize various online markdown to HTML converters or command-line tools.
Raycast Transform Extension: For Raycast users, the Transform Extension offers a "Markdown to HTML" command that can quickly convert copied Perplexity Page content to HTML.
While these methods provide workable solutions, they still require some manual effort.
As the demand for easier content export from Perplexity grows, it's possible that more user-friendly solutions or even native export options may be developed in the future. Until then, users can continue to leverage these workarounds or explore other emerging tools that may streamline the process of converting Perplexity Pages to website-ready HTML.
Social Sharing Image: Screenshot of the Source Perplexity Page
Source: Perplexity Page ➲ How BLUEOCEAN.law use Perplexity Pro AI?
Credits: This blog article was curated by James D. Ford Esq., GAICD CIPP/US CC | Attorney-at-Law, Blue Ocean Law Group using Perplexity Pro AI search result pages converted to HTML.
State of California Bar Number: 346590
Important Notice:
This blog article is intended for general interest + information only.
It is not legal advice, nor should it be relied upon or used as such.
We recommend you always consult a lawyer for legal advice specifically tailored to your needs & circumstances.
Listen to all of our Lawyer Logic Podcasts on Podcastle.ai here or on Apple Podcasts, Spotify or YouTube.
Katherine Hines
Mackay GoodwinHey James, I loved your article on perplexity! One of my good mates chewed my ear off for an hour talking about how amazing it is for his business, so it is great to see how it's being used in professional services!