The AI boom has brought an overwhelming number of tools, many offering similar features and capabilities. But not all tools are equally useful. This guide filters through the clutter to identify the AI platforms that stand out for research, writing and productivity.




The features and prices for these tools are constantly changing, so you should check the current plans before making a decision, but the fundamental strengths and trade-offs described here have remained rather stable.



How to Choose: Three questions before you pick




Before comparing tools, answer these:

• Which of these tasks takes the longest? Reading papers, drafting emails, summarising meetings, and organising projects involve different problems, each with its own best tool.


• How important is accuracy? Although you may afford to make errors during brainstorming, a school project or an office report requires a higher degree of accuracy.


• What job are you already doing? And which tools do you currently use, and what do you like and what do you dislike about them?

AI Tools for research




In research, AI can make the greatest time savings, but it is also in this area that it can cause the most harm if you place complete trust in it. The best research tools perform two functions: they enable you to locate relevant material more quickly than you could on your own, and they indicate the sources of their claims so that you can verify them.

• General-purpose research assistants include:

Claude, ChatGPT and Gemini all have deep-research modes which search the web, read a large number of sources and then provide a structured, cited report. They are very useful when you need to get a grasp of a subject you're unfamiliar with, when you want to compare different viewpoints, or when producing an initial briefing. Claude is generally commended for its careful reading of long documents and for its nuanced writing; ChatGPT for the range of its features and integrations; Gemini for its close integration with Google's ecosystem. In reality, most people keep one of them as their primary assistant and sometimes check the answers against those from a second one.




The system is based on search, since each answer includes inline citations and thus provides the quickest means of addressing a factual question while showing where the information originated from. It is a good alternative to the "search, open ten tabs, skim" approach, in particular when it comes to current events and product research.

• Academic and document research tools

Elicit and Consensus are intended for use in scholarly research; they search out literature in the academic field, extract the findings from the papers, and allow you to compare different studies. When you're carrying out a literature review, they can turn what would take days of browsing into just a few hours. With regard to questions such as "does X affect Y?", Consensus is especially useful since it summarises what the existing evidence indicates.




Semantic Scholar and Scite can assist with the next stage of research, which involves locating related papers and checking whether later studies have supported or contradicted a particular claim. The task of tracking how a finding has fared over time is one of the most useful things that AI can do for a researcher.




NotebookLM from Google allows you to upload your own materials (such as PDFs, notes, web pages, and transcripts) and then answers your questions solely on the basis of those documents, including citing the passages it has referenced. Since it works within your documents, it is much less likely to invent information. Furthermore, it is capable of producing study guides and audio-style summaries, something that is often useful to students and analysts when dealing with complicated material.

• A research rule that never changes

No matter what tool you use, you should check all the information that is important. AI systems are capable of misreading a paper, assigning a quote to the wrong source, or confidently summarising something that in fact doesn't exist. For every claim that you intend to rely on, open up the source that is cited. View AI as a highly efficient research assistant that at times may bluff, not as a definitive authority.



AI Tools for Writing




Writing tools can be divided into two categories: ones that produce text and others that help to improve text that you have already written. Choosing the right type for your needs avoids a great deal of disappointment.

• For drafting and thinking

Claude, ChatGPT, and Gemini form the basis of the majority of people's writing workflow. They are useful at all stages: helping to brainstorm possible approaches, producing outlines, drafting initial sections, rewriting for a different audience, and condensing overly wordy paragraphs. To make this process easier, a few prompts may be used.




Say to the tool who your audience is, what your objective is, and what tone you'd like. If you enter ‘Write a blog post about budgeting’, you'll get generic content; but if you specify ‘write a 600-word post for first-time freelancers explaining how to budget irregular income, in a friendly but practical tone’, you'll get useful material.




Put in a sample and ask the AI to copy the same voice. This is the most effective method of avoiding that generic ‘AI sound.’ Always consider the first output to be just a preliminary version. Instead of accepting the initial draft, ask for revisions, cuts, and other options.

• For polishing your writing and tone

Grammarly is still the most commonly used tool for picking up errors, modifying the tone, and making awkward sentences sound smoother when writing emails, documents and web pages. DeepL Write is a good alternative, in particular if you write in more than one language or are looking for clean and natural rephrasing. Both tools should be used after you have finished writing them, as a final check on the quality rather than in place of your own thinking.

• For writing inside your existing workflow

With Notion AI, Gemini in Google Docs, and Copilot in Microsoft Word, you can carry out drafting and summarising right within the programs you currently use. Although they are seldom the most powerful tool available, the convenience is important since it means you don’t have to switch between windows or copy text back and forth. Whenever you spend a lot of your day using one of these tools, the built-in assistant is usually the best place to begin.



AI Tools for increased productivity




The productivity AI sector is the one that is most saturated and also the one receiving the most hype; the tools worth paying attention to are the ones that get rid of repetitive tasks rather than presenting another dashboard for you to check.

• Meetings and notes

One of the best examples of AI is meeting notes. Tools such as Otter.ai, Fireflies.ai and Granola record or transcribe conversations and then generate summaries along with action points. Otter and Fireflies join calls as visible bots and are suitable for teams that want searchable archives; Granola, on the other hand, takes a more subtle approach by working with audio from your device and combining your personal notes with the transcript.

• Task and schedule management

Motion and Reclaim.ai make use of artificial intelligence to arrange your tasks in accordance with your calendar, automatically adjusting your day when meetings change. These services are suitable for individuals who have busy and constantly changing schedules and who end up spending a lot of time on manual planning. Although you might not need them if you already have a tool you trust, they can indeed be genuinely useful in the case of chaotic calendars.

• Automation

Zapier and Make allow you to link your apps and automate multi-step workflows, and both have now added AI steps which can summarise, classify or draft content as they go. For example, when a form is submitted, the AI step prepares a personalised reply, stores the details in a spreadsheet and sends you a message on Slack. Although the automations do require some initial setup, once you have created a good workflow, it can save you several hours each week.

• All-in-one workspaces

Workspace tools such as Notion are now bringing together notes, databases, project tracking and AI search in a single environment. The attraction of this is that you can ask questions about all the material you have saved—for example, ‘What decision did we make regarding the Q3 launch?’—and have the answer pulled from your own documents. The drawback is that the quality of these tools depends entirely on the information you enter.




The AI tools that are most suitable for research, writing and productivity are those which match the way you currently work; for the majority of people this involves having a single, powerful general-purpose assistant as a base, adding one specialised tool for the task which takes up the most of your time, incorporating a research tool such as Perplexity or Elicit, together with a meeting recorder or a scheduling assistant, and finally using a polishing tool such as Grammarly to pick up on anything you might have missed.




Choose a task, try out a tool for that task, check whether it actually saves you time or enhances your results, and only retain those tools which prove their worth. AI functions most effectively as a capable assistant that takes over the dull tasks, thereby freeing you to carry out the thinking, judging and creating that are uniquely yours.