• Everyday AI
  • Posts
  • The best AI tools to boost your research 🚀

The best AI tools to boost your research 🚀

AI and academic research, Dall-E 3 inpainting available to all, Stability's Stable Audio 2.0, a GPT for fewer hallucinations, and more!

Author

Everyday AI
April 03, 2024

Outsmart The Future

Today in Everyday AI
6 minute read

🎙 Daily Podcast Episode: AI is so underrated when it comes to research. With the right tools, AI can supercharge academic research. Here’s what you can equip yourself with for accurate AI research. Give it a listen.

🕵️‍♂️ Fresh Finds: An AI-powered healthcare bot, Google’s report on AI in developing countries, and has the AI gadget era arrived? Read on for Fresh Finds.

🗞 Byte Sized Daily AI News: Stability AI Releases Stable Audio 2.0, Microsoft Copilot for Business adds GPT-4 Turbo, and Replit’s Code Repair AI Assistant. For that and more, read on for Byte Sized News.

🚀 AI In 5: Wanna stop ChatGPT from lying? Here’s a GPT that reduces hallucinations. See it here

🧠 Learn & Leveraging AI: We’re breaking down how you can use AI for consistent and accurate research. Keep reading for that!

↩️ Don’t miss out: Did you miss our last newsletter? We talked about reasons why banning AI will lead to failure, US and UK partner on AI safety, new ChatGPT features Check it here!

AI Tools to Supercharge Research 🚀

An underutilized superpower of generative AI?

The way it can completely change how research is done.

So why isn’t it used more?

Because AI can spew out inaccurate information if not used correctly.

But if you have the right structure and tools in place, AI can do wonders for academic and scientific research.

 Avi Staiman, Founder of SciWriter.AI, joins us to discuss AI tools to supercharge your research.

Join the conversation and ask Jordan and Avi questions on AI and research here.

Also on the pod today:

• Responsible Use of AI ⚖️
• AI Tools for Research and Writing 🛠️
• Integrity and reliability of AI content 📝

It’ll be worth your 38 minutes:

Listen on our site:

Click to listen

Subscribe and listen on your favorite podcast platform

Listen on:

Here’s our favorite AI finds from across the web:

New AI Tool Spotlight – QuickBlox is an AI-powered healthcare bot, Storipress turns your blog into a lead gen engine, AI Talk is an AI language tutor.

Big Tech – Google has released a report on how AI can help grow developing countries.

AI Tech - With tech like the AI pin and Meta’s glasses, are we now in the AI gadget era?

Future of Work – This survey shows the struggles between employers, workers, and AI security.

Read This – Convenience store chain Re-Up, is adding a robot cook.

New Releases — OpenAI announced that their new inpainting-esque feature is now available in ChatGPT Plus.

1. Stability AI Releases Stable Audio 2.0 🔊

Stability AI's latest release, Stable Audio 20, allows users to create complete musical tracks up to 3 minutes long with coherent structures using latent diffusion technology. This innovative model marks a significant advancement from its predecessor, enabling users to produce immersive soundscapes and refine audio samples with natural language instructions.

2. Microsoft Copilot for Business adds GPT-4 Turbo and DALL-E 3 ✅

Microsoft is stepping up its game by offering priority access to GPT-4 Turbo and DALL-E 3 models for business subscribers of Copilot. With the latest AI models, users can enjoy no limits on chats, increased image generation capabilities, and enhanced text analysis across various work data.

3. Replit’s Code Repair AI Assistant 🧑‍💻️

Replit is transforming code repair with the introduction of a cutting-edge AI model tailored to the platform. This tool promises to enhance developers' efficiency by suggesting fixes and improvements in real-time as they code. Leveraging data from Replit sessions and LSP diagnostics, developers can now access a powerful Replit-native AI tool for efficient bug fixing.

4. US Military's Secret AI Playground Revealed 🇺🇸

Pentagon's Task Force Lima is creating a "virtual sandbox" for military personnel to experiment with generative AI tools, aiming to boost efficiency and effectiveness. Navy Capt. M. Xavier Lugo spilled the beans on this top-secret project, highlighting its potential to revolutionize military operations.

5. FDA Approves AI Tool to Detect Sepsis 🧑‍⚕️

Prenosis’ groundbreaking AI tool, Sepsis ImmunoScore, has received FDA approval after rigorous testing and validation processes. This innovative diagnostic tool utilizes 22 parameters to assess a patient’s risk of sepsis, a life-threatening condition often challenging to diagnose early.

A GPT for fewer hallucinations and better research!

Tired of ChatGPT feeding you inaccurate information?

We found a GPT that will give you more reliable information and fewer hallucinations.

Consensus is an academic AI research assistant and it’s a great GPT to use!

Or see this related video:

🦾How You Can Leverage:

If researchers are using AI to write research articles and we’re all using AI to summarize them, is it a never-ending AI loop? 

Should you use AI to summarize that research paper? 

Can we trust a research paper written with AI? 

And what tools should we be using to balance research speed and efficiency with depth and accuracy. 

Tough questions, we know. 🤔

Don’t worry — you know we took a deep dive and addressed them all on today’s show. 

We had an encore guest appearance from Avi Staiman, the founder of SciWriter.AI.

We broke down the love-hate relationship of AI in research, best practices to responsible research, and some AI tools that can supercharge research that even we didn’t know about. 

Buckle up. 

Because if you’re in the research space or you’re just looking to have a conversation with that scientific paper to figure out what that rash on your arm is, then today’s show can be both your starting point and your guide forward. 

So, what do you need to know? 

Let’s get scientific wit it 🔬

1 – The price tag of research 🤑

One question we all might have is: why do we even need to use AI in research? 

Well, apparently it’s pretty costly to have established university minds create the research our world relies on. 

Avi mentioned that the average cost of producing a single research paper can be more than $500K.  

Yikes. 😳

That figure includes not just the direct expenses of writing and publishing the paper but also encompasses the broader infrastructure required for conducting the research. 

So naturally, research institutions have turned to AI to not only reduce the staggering costs, but to also create research faster so humanity can benefit. 

Try this:
This Heritage Foundation study looks at the rising costs of research papers at public institutions and the burden that taxpayers $ometimes $houlder.  

2 – The love-hate relationship with AI 😍 

You know what researchers pride themselves on? 

Being factual. 

You know what sneaks up on researchers who rely too heavily on AI? 

Falsehoods. 

Avi talked about the research industry’s love-hate relationship with AI. 

Because researchers are definitely using AI to accelerate their discoveries, but sometimes it leads to hallucinations from Large Language Models making their way into academic research. 

Or, laughingly large rat testicles (generated by AI) making their way into scientific journals. 

So how can the research industry balance AI-powered research with making sure little LLM lies don’t make their way to the internet? 

(And ughh…. Future large language models that gobble up the Internet?) 

Avi said researchers need to ‘run the rational middle path’ between efficiency and accuracy. 

Try this: 
How is that even possible?

This article from Scholarly Kitchen gives great advice on how to strike the balance Avi talked about between using AI for research and keeping it factual. 

3 – Tools to supercharge the process ⚡

Oh, you like tools? 

(Same shorties.) 

Here’s the quick facts on AI tools that can help you find more reliable sources of information that you can actually cite and rely on. 

First things first — Large Language Models out of the box are traditionally pretty bad at this. (Mainly because of their knowledge cutoffs and inability to use the internet correctly.) 

ChatGPT did just improve its sourcing and citations this week, which we covered here. 

But, there’s still an increased need to leverage GenAI in a responsible way, which often includes using dedicated services for AI-powered research. 

Try this:
Avi didn’t disappoint with his recommendations for AI tools that can reliably help in the research process. (That are also great for non-researchers who want to more quickly understand research papers!)

Some of his (and our!) favorites: 

ScholarAI (we did an episode with them here!) 

Tool-Up Tuesdays (From Avi)  

⌚

Numbers to watch

22%

22% of North American employees admit to violating company rules on the use of generative AI.

We wanna hear from you below!

What are your thoughts on our Numbers to Watch section?

Login or Subscribe to participate in polls.

Reply

or to participate.