As with all things in a GIGO newsletter, we strive to bring you meaningful updates to advances in the world of programming. Sticking with the common theme of advances in AI, some engineers are releasing not only creative applications, but genuinely useful tools that can be leveraged in the day to day tasks of your average white collar worker.
In this edition, we are thrilled to dive into the fascinating world of Perplexity AI, a cutting-edge technology that is reshaping the landscape of searching for information online.
The days of tearing thru books in the library and physical research papers is mostly phased out, but the days of sorting thru practically infinite pages on the internet for your answers is in full stride.
Having an AI application to sort the mess and deliver your desired key points to the top of the pile will soon be looked at as a necessity.
In short, Perplexity is the worlds first conversational search engine. But wait, why do we need a new search engine? What about the almighty Google whom we as developers trust and accredit so much of our success to? Google has done a fantastic job aggregating their list of websites and making sure that what you’re looking for comes to the top of the list when you tip tap & launch your little characters into the search bar. But in the modern skilled workers world, this has simply become not enough.
Instead of entering keywords and sorting through a tangle of links, users pose their questions directly to Perplexity.ai and receive concise, accurate answers backed up by a curated set of sources.
Perplexity Performance Achievement
The organization hit the ground running in August 2022 led by co-founders Aravind Srinivas (CEO), Denis Yarats (CTO), Johny Ho (CSO) and Andy Konwinski.
Their mission is to build the world’s best answer engine — one that people trust to discover and expand their knowledge. To achieve this, they are deeply focused on providing helpful, factual, and up-to-date information.
To benchmark their LLMs’ performance on these axes, they curated evaluation datasets to reflect challenging yet realistic use cases for answer engines. For each query in each evaluation set, contractors were given two model responses and instructed to select the response that performed better for the following criteria:
Helpfulness: which response answers the query and follow the specified instructions better?
Factuality: which response provides more accurate answers without hallucinations, even for questions that require very precise or niche knowledge?
Freshness (inspired by FreshLLMs): which response contains more up-to-date information? A model excels in this criterion if it is able to answer queries with “fresh” information.
In addition to the three criteria above, model responses were also evaluated holistically. To evaluate responses holistically, evaluators were asked to pick the response they would rather receive from an human assistant who is helping with the query.
Perplexity recently released two new PPLX models: pplx-7b-online and pplx-70b-online! Their online models are focused on delivering helpful, up-to-date, and factual responses, and are publicly available via pplx-api, making it a first-of-its-kind API. pplx-7b-online and pplx-70b-online are also accessible via Perplexity Labs, their LLM playground.
LLMs have transformed the way we find information. However, there are two limitations with most LLMs today:
- Freshness: LLMs often struggle to share up-to-date information.
- Hallucinations: LLMs can also output inaccurate statements.
Their pplx-7b-online and pplx-70b-online models address current limitations by additionally providing helpful, factual, and up-to-date information in its responses.
Back to the Future
A week ago (Jan-4th-2024) Perplexity announced that it raised $73.6 million in a funding round led by IVP with additional investments from NEA, Databricks Ventures, former Twitter VP Elad Gil, Shopify CEO Tobi Lutke, ex-GitHub CEO Nat Friedman and Vercel founder Guillermo Rauch. Other participants in the round included Nvidia and Jeff Bezos.
This valuation places the up and coming search engine at a roughly $520 million valuation. Considering that Perplexity’s only been around since August 2022, it’s a significantly impressive feat and signals a journey worth keeping an eye on.
To be added to the GIGO Tech Trends weekly Newsletter mailing list please reach out directly to us on this account or email us at contact@gigo.dev
Discord: https://discord.com/invite/learnprogramming
Twitter: https://twitter.com/gigo_dev
Reddit: https://www.reddit.com/r/gigodev/
GitHub: https://github.com/Gage-Technologies/gigo.dev
Get started on your first project at gigo.dev
Top comments (0)