Get inspired by AI and understand which tools have the potential to give you the most leverage in your workflow, explains Fiona Young
Everyone’s talking about ChatGPT, and for a good reason. AI chatbots like ChatGPT, Copilot, Claude, and Gemini are incredibly powerful tools.
But there’s a whole universe of purpose-built AI tools for work emerging beyond these.
That’s what I’ll cover here:
- An overview of the AI tools most useful for the Assistant workflow
- A deeper dive into 3 AI tools forward-thinking Assistants are using to shortcut admin right now
- How you can find, vet, and propose AI tools at work, with practical tips to overcome barriers you may come up against
What Is Gen AI?
Before I get into tools, a little background on AI. When you hear the term ‘artificial intelligence,’ the first image that springs to mind is most probably robots, sci-fi, and autonomous vehicles.
Yes, these are AI, but the reality is that AI is also a lot of the mundane things that have been around for decades. Think spam filters. Or credit card fraud protection that will freeze your card with any unusual transactions. Or recommendation engines in Spotify or Netflix.
What’s new is the development of generative AI or ‘gen AI.’ This subset of AI describes a family of models that are all about creation. They can autonomously create completely new content like art, music, text, video, voice, and code.
The AI productivity tools I’m covering in this article use the power of gen AI to shortcut work.
A Snapshot of the AI Tools Landscape Today
Many of the productivity tools for work use the power of large language models (LLMs) to create human-like work outputs. LLMs are a type of gen AI trained on a huge amount of text data that can understand and generate text in a human-like way. The engines behind AI chatbots like ChatGPT are LLMs. For example, LLMs are used in slide deck tools to generate the written content in a slide deck, or in a meeting tool to generate a transcript and distil that into a meeting summary.
The dirty little secret of AI tools for work right now is that many of them are simply ChatGPT reskinned. Developers have taken GPT-4 (the LLM behind ChatGPT) and integrated it into their product, adding an invisible layer of nuanced instructions and a very polished UX (user experience) on top.
One particularly well-funded example of this is Jasper for content writing. It’s little more than ChatGPT, but easier to use and easier to get great written outputs from – so people are willing to pay USD 60/month for it.
It is worth understanding AI chatbots, as there’s a lot you can do using these alone when you’ve learned the art and science of prompt crafting (writing strong questions or instructions to put into the chatbot).
There’s also an interesting dynamic playing out between independent versus native tools. (By native tools, I mean Copilot built into the M365 suite and Gemini built into Google suite products.)
At the moment the capabilities of independent tools are well ahead of Microsoft and Google. But over time I expect they will catch up, and there will be significant benefits to using these native tools for both convenience and security.
The Top Categories of Gen AI Tools Most Useful for Assistants
Meetings
AI meeting notetakers join your virtual meeting, listen in on conversations, create a full transcript, and write up a summary and actions. Some have a ‘catch me up’ feature to summarize what was said if you join a meeting late. And some will also give you real-time feedback on your performance, flagging when you’re monologuing or when the AI tool senses others are tuning out when you’re speaking.
AI email tools now write and summarize emails for you. The best ones train on your past emails and can mimic your tone, style, and commonly used phrases to sound exactly like you.
Calendar
Advanced calendar tools allow you to create events using natural language by typing “coffee Jill Baker 3 pm Friday Blue Bottle SoMa” rather than having to add in each of those parameters manually. I expect there will be more AI features for strategic calendaring available by late 2024, such as slots suggested within your draft email replies using available data that can indicate urgency and importance (for example, if the other party is internal versus external, or how urgently you scheduled them in the calendar last time).
Task and project management
These tools can use a simple prompt from you to create highly detailed project plans and task lists and prioritize tasks for you based on your deadline. Some AI tools act as a planner, presenting tasks and calendars side by side and slotting tasks into your calendar where you have gaps (great for those who love timeboxing tasks).
Slide decks
Though I don’t think any of the AI slide deck tools are truly capable of creating decks as well as humans can yet, they can create a whole presentation from a single prompt, including presentation structure, AI-generated imagery, written content on slides, and slide notes. But for now, be prepared to do some heavy editing to get your deck to a high standard!
Standard operating procedures (SOPs) and processes
If you often need to document processes or create how-to guides in your organization, these tools are a huge time saver. You switch them on and proceed to do the thing you need to create a how-to guide for, and the tool ‘watches’ your mouse clicks and instantly generates a detailed guide for you, including video snippets if you want them.
3 AI Tools Assistants Are Loving Right Now
Scribe: for creating SOPs and how-to guides
You can use Scribe in a browser or on your desktop. When you switch Scribe on, it watches your mouse clicks and instantly generates a step-by-step guide for you. You can manually add steps or tips if you need to and share a link with colleagues.
Taskade: for task and project management
If you do a lot of project work, Taskade might be the tool for you. It creates highly detailed project plans and task lists from a simple prompt and has AI agents built in that can act as an expert and do some of your project work for you, like research and writing.
Read.ai: for meeting notes and insights
Read.ai joins your meetings, records them, and creates a summary and draws out actions for each. It also gives you insights on things like how much you talked and whether it detected any bias in your behaviour in the meeting. You can choose to have Read.ai’s summaries automatically sent to meeting attendees afterwards.
How to Find, Vet, and Propose Tools at Work
It can be daunting to try to figure out what to use given just how vast the AI tools landscape is.
Identify your biggest time sinks
The best way to start is toidentify your biggest time sinks:
- What tasks do you pour hours into every week?
- What tasks do you frankly hate doing and find yourself procrastinating doing every week?
Search for a tool that can help you with those. There’s a dedicated tool out there for just about everything you can imagine.
Discover vetted tools
I’ve put together a free guide with my favourite AI tools for Assistants. It contains over 50 productivity tools, detailed information on their features and pricing, and honest reviews based on hours of testing. Use the QR code to get your copy now.
I also recommend the website There’s an AI For That (theresanaiforthat.com), with the caveat that it can be a little overwhelming. It’s a giant, searchable repository of 12,000+ AI tools (and counting).
Proposing AI tools internally
So you’ve found a few tools that you want to try. The next step is to get approval from your organization, usually your IT or infosec team.
I don’t recommend incorporating an AI tool into your workflow until you have express approval to use it.
Most AI tools access sensitive data like the contents of your emails or your executive’s calendar, so it’s important to make sure the tools will store and use that data safely.
To help with getting approval, do your own research first on the tool.
Some important things to keep in mind
- Read the terms: Make sure you know how the AI tool will store and use your data.
- Check their compliance with risk management standards, like ISO 31000 in the UK or SOC2 in the US.
- Be wary of free trials: Many AI tools offer free trials or limited free use; check your company policy, as some companies have rules against this (often when you’re using a free trial you have limited data protections).
4 Top Tips to Get New Tools Approved
1. Make the business case
Get really clear on the benefits of the tool to you and your team, using specific use cases. Do a rough ROI calculation, comparing the money saved each month (considering your effective hourly rate and the hours it will save you) vs the cost of the tool. Include links to details on security and compliance that speak to IT’s potential concerns.
2. Suggest starting small with a pilot
A small, timebound trial with you and one or two others will minimize risk and give your team the chance to road test the tool over a longer period. This is useful so you can get some real-world data on the value of the tool, and spot early if there are any issues with it.
3. Rally supporters internally
Get your exec and your colleagues on side. Take a quick poll on how many people are interested in getting access to the tool – there’s strength in numbers (and cost savings too).
4. Persist!
Don’t give up if the first response is ‘no’! Take time to figure out why your IT team rejected it, and the reasoning behind it so you can see if it’s possible to find a workaround. It’s not unusual for Assistants to carefully negotiate new tools with IT over the course of 2-3+ meetings.
One Final Tip
Recognize it will take time.
We all have deeply entrenched habits with how we work. Changing them takes time. Understand that you aren’t going to change everything overnight.
The point is to get your feet wet – get inspired by the potential of AI, understand which tools have the potential to give you the most leverage in your workflow, and set a goal to find 1-2 AI tools this year that will help you shortcut admin drudgery.
