No, not the board game. Let's start with something frivolous, for the sake of amusement. If there's one subject that could do with lightening up it's politics, so let's give that go.

I have taken the Wikipedia page on liberalism and asked it to explain it in the voice of Marvin, the depressed robot from The Hitchhiker's Guide to the Galaxy.

TRY THIS OUT YOURSELF²: in ChatGPT or Claude:

Explain conservatism in the voice of William Shatner

AI is quite good at summarizing text. Something computers were notoriously bad at just a few years ago, so the bar was low.

These kind of closed data sets are often where GenAI shines because you almost completely eliminate hallucination.

TRY OUT²: Summarize Documents Quickly

• Paste a document in to ChatGPT, Claude or similar, and enter the prompt:

Summarize the following text in fewer than 200 words, preserving key points.

Use case: If you’re dealing with lengthy work documents, feed them into an AI summarizer and use its output as a rough draft. Then revise.

Let's take our text from Marvin and put it through the mill:

It seems to have preserved some of the sentiment alright.

A notable mention here is Google's NotebookLM, where you can upload a collection of documents and interrogate it with questions. It can even generate a podcast for your ears, so that stack of boring policy documents you need to digest before the meeting can be packaged up for listening on your morning commute. It can make the mundane seem almost interesting. But use with caution as there are real concerns on the privacy of Google's AI services versus the others.

Generative AI has become so good at writing prose that OpenAI abandoned its project in detecting AI-generated content³. But there's still that niggling problem of hallucination (the technical term used to describe making things up). Often downplayed with the retort "humans aren't 100% accurate either". No, but the difference is the AI doesn't know when it's made something up. It sounds just as confident and can be quite persuasive – if you didn't know, you wouldn't catch it. In a world where disinformation is already weaponized, automating the creation of alternative facts is a real concern.

TRY OUT² with ChatGPT or Claude:

Help me create a bullet-point outline for an article on the history of tea in England. Emphasize key turning points, influential figures, and fun cultural anecdotes.

Benefit: If you know your subject, but struggle to get started on your outline, this saves time in brainstorming, and you can further refine this through more specific follow-up instructions. If you are winging it, this could be a great starting point for research, but you need to be wary that perhaps it has missed something important, or worse, it has made something up. In that case you need to verify everything.

TRY OUT² with ChatGPT or Claude:

Make this email more formal and diplomatic while preserving the core message

Benefit: AI can shift tone and style quickly, helpful for switching styles between a business email, or a lighthearted blog post.

As well as helping brainstorm and scaffold your writing, these tools are democratising the ability to write well for those who are less confident, or who may not be communicating in their native language.

ADDITIONAL MENTION: GRAMMARLY

I mention Grammarly because of its widespread use particularly among non-native speakers writing professionally in English. It is a tool that has been around for a while to help improve your writing, and can run inside many tools such as Microsoft Word among others. It is not without its problems. If you are a fluent, serious writer it will probably get in the way more than it helps, but as a product it is rapidly incorporating GenAI into its feature set. While often used by teachers to detect plagiarism - rapidly becoming redundant with the advent of this new set of AI technologies, it now offers to write articles for its students, which it can't detect as plagiarism. Go figure. I don't use this tool much but I probably con't do better than this review from PC Magazine.

There will be plenty of fools who ask AI to write the story for them. And failure to disclose it could risk hefty fines⁴. But for more serious writers, there are techniques in helping with the brainstorming process that are good to know.

TRY OUT² with ChatGPT or Claude:

Give me 10 unique characters who live in a remote Irish coastal village, with potential secrets or emotional baggage.

7 of these will probably be garbage. But 3 might be interesting and something to build on.

TRY OUT²:

Take characters #1, #3 and #7. Now give me 5 inciting incidents that would force these characters to interact in uncomfortable ways.

Again, most of these are probably rubbish, but one or two might be something worth building on. And further interrogation of the AI model through follow-up prompts could serve as a useful brainstorming exercise.

So far we have been looking at the dry interface of a text box chat bot. We have yet to see this tech enter the world of voice assistants but it is coming. A decade ago the likes of Alexa and Siri were novelties that for a moment we thought might change the world, but these days have been relegated to kitchen timers. Though in the world of accessibility, it has been transformative for people like me with a disability, enabling me to communicate and read hundreds of (audio)books.

I highly recommend trying out the voice interface on ChatGPT's mobile app (Apple iOS or Android). Ask a few questions about any topic - and follow up questions. The difference in capability versus the current generation of Alexa or Siri is astonishing - I urge you to download the app and play with it. This level of capability is coming to your iOS, Android and Alexa devices soon.

I keep hearing about it's poetry writing skills. Sure, it can rhyme words and count syllables, but I challenge anyone to have it generate a poem, and then go read it aloud in front of another human. Personally I'd be mortified to be caught even reading it.

^ This is what I wrote in my first draft of this article, which I sat on for a few weeks. Then by chance in my podcast feed, in the most unexpected of quarters, Richard Dawkins challenged ChatGPT to write a sonnet praising biological evolution, read it aloud then critiqued it. He misses the point slighlty on his surprise at it being non-deterministic⁵. But he seems to think the sonnets are passable, with a few human corrections required.

Which leads me to challenge my own premise. As these models improve - if it eventually does start to churn out poetry that's better than its currently passable efforts, I mean, so what? Computers surpassed humans at chess decades ago. Did we stop playing chess? Did chess enthusiasts stop following the grand masters? Not one bit.

The more recent example of this is the game Go, which was thought to be too complex for a computer to master⁶. But it did, and it produced techniques so unexpected that it changed the way humans play the game. Maybe some day it will do something interesting with poetry.

I won't reproduce the exercise here, instead I will point you to Dawkins experiment and his conclusions. It's a fun listen.

This could be an extensive piece on its own. I have a project in the works in this area for which I will publish more details on this blog soon. In the meantime I will try and give an overview of AI developments in accessibility. There may be a slight bias to areas I have personal experience.

If you will excuse a small detour into my personal story, I have quadriplegia – well, incomplete quadriplegia to be precise. In July 2022 an injury rendered me paralyzed, with no movement or sensation below the neck. For six months I stared at a hospital ceiling immobilized. The doctors were telling my family to lower our expectations, that I would need 24-hour care for the rest of my life. Voice interfaces in both consuming podcasts and audiobooks, and interacting with my phone and computer, saved my sanity.

Somehow, through pig headed stubbornness, and an excruciating three years of physical rehabilitation I have been extraordinarily lucky in recovering some mobility and use of my hands again. But this is rare. Most people with paralysis do not recover this level of function. My cervical spine is held together with metalwork, which has acquired a large fracture since it was installed. It weighs on my mind that some day I may lose these gains, and return to full paralysis.

For this reason I have a very keen and vested interest in technologies that enable interaction with technology in both hands-free and low-friction mechanisms.

Speech recognition

Not just for setting kitchen timers. AI has changed voice recognition from a rigid interface, where you had to speak in a rather robotic way to be understood. It is now a realistic method for

• controlling your computer, smartphone or tablet
• dictation for text input, with good controls for editing text that were previously clunky and awkward

Under the hood, modern MacOS computers and iOS phones and tablets have been quietly improving the fidelity of voice input recognition and intents, which has opened up a level of dexterity in what used to be an awkward, stilted mechanism. While Windows and Android are catching up, they are not quite there yet, but additional software such as Dragon NaturallySpeaking and Dragon Nuance open up these platforms to the same level of control.

And then there are community projects such as Talon, which is a voice control and dictation software, as well as eye tracking system, that works on MacOS, Windows and Linux. It is designed for people with disabilities, but is also used by gamers and programmers with RSI⁷. It has a steep learning curve, but once you get the hang of it, it is incredibly powerful.

I will start however with the built-in tools in MacOS and iOS, simply because they are free⁸, easy to use, and already in so many people's hands.

TRY OUT²:

• On MacOS, go to System Preferences > Accessibility > Voice Control and turn it on. You can now control your Mac with your voice.
• On iOS, go to Settings > Accessibility > Voice Control and turn it on. You can now control your iPhone or iPad with your voice.
• For other platforms, take a look at Dragon sofware options including free trial versions.

In iOS or MacOS, the first thing you should do is open the Voice Control Tutorial (or Guide as it's called in MacOS).

iOS voice control tutorial ▼

macOS voice control guide ▼

Low-friction typing

Text replacement is a feature that has been around for a while, but it is now being used in new and interesting ways. It allows you to create shortcuts for frequently used phrases, sentences, or even entire paragraphs. This can be particularly useful for people with disabilities who prefer to use a keyboard but still find it a slow process.

macOS and iOS have built-in text replacement features, but enough Apple already. TextExpander has been the industry leader in this space for years. However my personal favourite is Espanso, because it is privacy-first, free, open-source and cross-platform.

It works by setting short codes or phrases that automatically expand while typing. For example, typing "addr" could expand to your full address, or "sig" could expand to your email signature. You can also set up more complex snippets that include variables, such as the current date or time. Espanso also lets you search for snippets using keywords, so if you can't remember a shortcode, so you can quickly find the one you need.

Another interesting one is Amazon Q, which is intended for programming, but also works for prose. Since I write for my blog using marked up text in VS Code, Q proposes my next sequence of words as I type. Most of the time it's off the mark, but as it learns my writing style and context, the further I get in to a sentence, the more accurate it becomes at predicting my next few words. In the following example, it saved me typing 5 words. However, it is debatable how much time it actually saved me since I had to delete everything after the comma:

I show this rather imperfect example to illustrate how fast things are moving – while not part of the suite of text replacement tools directly, yet, accessibility tech is in a very exciting place where things are improving, and improving people's lives, very quickly indeed.

Visually impaired

Some of these are not new but are being augmented by AI. I include them here because they are so simple yet I'm constantly surprised at how many people struggle to read their devices when help is just a few clicks away.

• Zoom: All OS's⁹ have built-in zoom features that make it easy to quickly enlarge text on the fly then return to a normal layout screen. This one means you don't have to sacrifice screen real estate in navigating your device.
• Contrast: All OS's have built-in contrast features that make it easy to quickly change the contrast of text on the fly. This one means you don't have to sacrifice screen real estate in navigating your device.

• Filters for colour-blindness: All OS's have built-in filters for colour-blindness. They filter the blind colour, and give it a higher contrast in a visible colour to the user so that the semantics of the colour are preserved. All platforms cater for:

  • Protanopia (red-blind): Boosts reds and differentiates them from greens.
  • Deuteranopia (green-blind): Boosts greens and makes them stand apart from reds.
  • Tritanopia (blue-yellow blind): Alters blue and yellow hues to increase contrast.

In addition macOS and iOS have further options

• Colour Tint (MacOS/iOS): A custom tint overlay that you can adjust manually.
• Greyscale (macOS/iOS): For users who are fully colour blind (achromatopsia) or prefer no colour.

Example zoom controls ▼

Enabling high contrast ▼

Example colour filters ▼

So far its ability to parse natural (human) language that has been the key development of LLMs. So it should not surprise us that it can in turn write computer code – programming languages are more formalized in their rules and structure. At least, it can generate code until something doesn't compile, and if you don't know the basics you can get in to trouble. I've been using it in my own programming for a while now and it frequently makes up a non-existant API call¹⁰ or function. The old hallucination problem again.

However it really shines, I think, in helping humans both learn and write code.

I made a little open source project a while back to generate Kriskindle matches (a.k.a. Secret Santa) at a certain time of the year.

TRY OUT²: In your IDE¹¹ install the plugin for GitHub's Copilot or Amazon Q. (Works with VS Code but also JetBrains, Emacs, even Vim)

Here's the source code.

Explain this code, what is it doing, and how does it work?

# Input format example:
# Bob Jane Mary    # Family group 1
# John Sarah Tim   # Family group 2

# Example output:
# Bob → Sarah
# Jane → Tim
# Mary → John

It's actually done a pretty good job of explaining my code.

Let's write a small program. When learning a new language, the traditional first exercise is to print Hello World to the screen.

TRY OUT²: In your IDE¹¹ with GitHub Copilot or Amazon Q plugin, create a new empty file, and save it as hello-world.sh.

Now open the Amazon Q or Copilot chat window, and prompt:

Help me write a simple shell script that prints  Hello world  to the screen

Not only has it given us the starting code, it has explained how to make the script executable¹².

Now, in our Hello World program, let's add a comment¹³:

# now let's print numbers from 1 to 100

And our Autocomplete prompt, in this case generated by Amazon Q, has suggested the next block of code, based on the intent we have declared in our comment.

Now let's try something trickier and ask it to create a whole program:

Write a Python script that reads a CSV file of customer transactions and outputs the top 5 customers by total spend. Include comments explaining each step.

import csv
from collections import defaultdict

# Define the path to your CSV file
csv_file_path = 'transactions.csv'

# Dictionary to store total spend per customer
# Using defaultdict so we don't need to check if key exists before adding
customer_spend = defaultdict(float)

# Open the CSV file for reading
with open(csv_file_path, mode='r', newline='', encoding='utf-8') as csvfile:
    reader = csv.DictReader(csvfile)
    
    # Loop over each row in the CSV
    for row in reader:
        customer = row['customer_id']  # Assumes CSV has a column named 'customer_id'
        amount = float(row['amount'])  # Assumes CSV has a column named 'amount'
        
        # Add the amount to the customer's total spend
        customer_spend[customer] += amount

# Sort the customers by total spend in descending order
sorted_customers = sorted(customer_spend.items(), key=lambda x: x[1], reverse=True)

# Print the top 5 customers
print("Top 5 Customers by Total Spend:")
for customer_id, total in sorted_customers[:5]:
    print(f"Customer ID: {customer_id}, Total Spend: €{total:.2f}")

It's a pretty good solution, and GenAI is fast becoming an essential tool in every programmer's war chest to writing better code and faster. Its autocomplete capability can now suggest entire blocks of code which are unnervingly close to reading your mind.

• Good: Just copying and pasting text to and from ChatGPT, it can answer questions and generate code quite capably.
• Better: GitHub's Copilot is well worth the $10 per month if for its inline suggestions alone. However there are question marks over whether and how it might store your code as a personal user².
• Best: Amazon Q, which is just as capable as Copilot, except that it's free for personal use, and does not store or harvest your code.t

The capabilities go far beyond these simple examples I have shown above. It is capable of generating a smartphone app, or an entire website.

The biggest danger I foresee is idle talk of AI replacing programmers. From everything I've seen, it is changing the way we code, but the career of the developer is not going away any time soon. There is already a great shortage of people coming in to this line of work, and the idea that it is going away only serves to put young people off choosing it as a career path, leading to a more chronic shortage in the industry.

• I examined this topic here – verdict: B-minus

And there are wonderful developments in this area helping blind and visually imapired people I mentioned in the Blind users (Accessibility) section.

Using GenAI in day to day research is fraught with many pitfalls, especially if you choose the wrong tools. Chat interfaces like Claude and ChatGPT have knowledge based on a fixed date on which their respective models were trained¹⁴.

However there are particular services emerging that has Google seriously worried for its monopoly¹⁵ share in the search market.

Two in particular stand out

• Perplexity AI is a search engine with a genAI model on top. It will parse the results of your search, summarize the results and attempt to answer your question directly in plain English¹⁶.
• Arc Search (mobile version only) Apple iOS or Android will do the same, but as part of the browser rather than a standalone website.

Good use of AI

Try out² using Perplexity or Arc Search

how do the recent changes in regulation affect me as a first time house buyer in ireland?

Perplexity AI full answer ▼

Recent regulatory changes in Ireland significantly impact first-time buyers through enhanced affordability schemes, relaxed planning rules, and accelerated housing delivery plans. Here’s how these changes affect you:

---

1. Extended Help to Buy Scheme (HTB)

The HTB scheme has been extended until 2029[3][11], offering a tax rebate of up to €30,000 (or 10% of the property price) for new homes or self-builds valued under €500,000[9][11]. This reduces upfront costs, especially when combined with a 10% deposit requirement[4].

• Eligibility: Must be a first-time buyer purchasing a newly built home or self-build, with a mortgage covering at least 70% of the price[4][9].
• Impact: Lowers deposit burden, particularly for buyers targeting new developments[3][10].

---

2. First Home Scheme Expansion

This shared-equity scheme allows buyers to borrow up to 30% of the property price (capped at €500,000) at 0% interest for the first five years[7].

• Eligibility: Open to first-time buyers and “fresh start” applicants (e.g., divorcees or those exiting insolvency)[7].
• Impact: Reduces mortgage size, making higher-priced homes more accessible[7][10].

---

3. Increased Mortgage Borrowing Capacity

Since 2023, first-time buyers can borrow up to 4× their annual income (previously 3.5×)[4]. For example:

Income	Borrowing Capacity
€40,000	€160,000
€65,000 (joint)	€260,000

This expands purchasing power in a market where the median home price is €359,999[8].

---

4. Planning Regulation Changes

• Detached Cabins: Proposed exemptions for modular homes (e.g., granny flats) in back gardens could provide affordable temporary housing options while saving for a primary residence[1].
• Faster Approvals: The Planning and Development Act 2024 streamlines processes, with local development plans now valid for 10 years (previously 6) and quicker judicial reviews[6].
• Supply Boost: The revised National Planning Framework prioritizes housing near public transport, aiming to accelerate delivery of 40,000+ annual homes[2][10].

---

5. Market Challenges

• Prices: Modest increases expected due to supply-demand imbalance, especially in urban areas[10][8].
• Interest Rates: Potential rate hikes may reduce affordability, offset partially by HTB and shared-equity schemes[10].
• Land Registry Delays: Slow processing times could prolong closing periods, requiring careful planning[8].

---

Key Takeaways

• Act early: HTB and First Home Scheme benefits are most effective when paired with new builds.
• Target designated zones: Focus on areas with Urban Development Zones for better infrastructure and supply[6].
• Consult lenders: Optimize mortgage applications using the 4× income rule and shared-equity options[4][7].

These changes collectively improve affordability and supply-side flexibility, though market pressures remain. First-time buyers should leverage state supports while monitoring interest rate trends[10][8].

Citations:
[1] https://ecosmarthubs.ie/exemptions-for-detached-cabins/
[2] https://www.gov.ie/en/press-release/60a7d-government-approves-revised-national-planning-framework/
[3] https://homsassist.ie/news/budget-2025-help-to-buy-scheme-key-changes/
[4] https://www.axis.construction/news/first-time-buyers-ultimate-guide-ireland-2023
[5] https://www.firsthomescheme.ie/faqs/rules-and-eligibility/
[6] https://www.abacuslegal.ie/planning-law-in-ireland-changes.html
[7] https://www.citizensinformation.ie/en/housing/owning-a-home/help-with-buying-a-home/first-home-scheme/
[8] https://homsassist.ie/news/understanding-the-irish-property-market-the-impact-of-land-registry-delays/
[9] https://www.citizensinformation.ie/en/housing/owning-a-home/help-with-buying-a-home/help-to-buy-scheme/
[10] https://mccarthy.ie/blog/house-purchase-sales/forecasts-for-irelands-housing-market-in-2025/
[11] https://www.sherryfitz.ie/news/Budget-2025
[12] https://www.oireachtas.ie/en/debates/question/2025-04-03/244/
[13] https://www.oireachtas.ie/en/debates/question/2025-04-01/478/
[14] https://www.mccannfitzgerald.com/knowledge/real-estate/whats-in-store-for-the-real-estate-sector-under-the-programme-for-government-2025
[15] https://www.citywidefinancial.ie/post/applying-for-a-mortgage-in-2025-homebuyers-guide
[16] https://www.revenue.ie/en/property/help-to-buy-incentive/index.aspx
[17] https://www.rte.ie/news/2025/0219/1497554-housing-planning-laws/

---

Answer from Perplexity: pplx.ai/share

This is a good use of AI simply because it has done an up to date real time¹⁷ search of the web, summarized the results - but crucially, it has given us citations with real sources that can be checked. Of course, that's not the same thing as a fact-check, since any fool can put any old sh*t up on the internet (ahem :), and the internet is where it is sourcing its answers. But here you at least have sources that can be checked.

Behind the scenes, this is powered by a search using Bing, finally finding a use for the much unloved alternative to Google.

When do the clocks change in Ireland?

This one has been around a bit longer and is better than it's ever been. In reducing the friction of communication, a fantastic resource. But would you rely on an AI-translated document, without any human scrutiny, for a legal contract? A law? A medical diagnosis? I was recently involved in the translation of a children's book, and it reinforced my view that it's a great accelerator, but much time and care was still needed to shape the language before it was ready to put in to the hands of a child.

How good are they? Take a look at some interesting results as I put Chinese poet Li Po's poem Lament of the Frontier Guard through various translation engines and then compared with the very famous Ezra Pound translation.

TRY OUT²:

• Deepl is a paid online service with a free tier and is also available as a plugin for Microsoft Word and Google Docs. It does not store the documents you upload for translation.
• LibreTranslate is a free and open-source translation tool that can be installed on your own computer. It is a great alternative to the online services and the various privacy concerns that can go with them. Running this on your own computer locally means you can guarantee your own data security.
• Google Translate is the most well-known, but it is also the most privacy invasive. However, for day-to-day use while travelling, it is hard to beat the Lens function that lets you see translation of text on, say, a food packat or road sign. Equally its voice/audio interface can enable communication between people who don't speak the same language in real time.

This may seem like I am coming down very negatively on it. To be clear - what generative AI can do in 2025 is nothing short of astounding. I am trying to articulate some of the caveats we will all need to be aware of as it becomes part of our lives.

KNEEJERK REACTIONS In the wake of powerful tools like ChatGPT, Gemini, and Claude, many schools and universities have rushed to ban or heavily restrict AI in the classroom. These bans often stem from fears that students will rely too heavily on AI to complete assignments or lose out on honing essential critical-thinking skills. Recent examples, particularly across Ireland and Europe, highlight the breadth of these concerns:

• Italian Ban on ChatGPT: Italy’s data protection authority briefly blocked ChatGPT over privacy worries, forcing schools and universities into a sudden scramble to figure out whether they could use AI tools at all.
• Irish Universities Debating Bans: Some Irish universities reviewed whether to block access to large-language-model websites on campus networks, concerned about academic integrity and reliability of AI-generated content.

These instant lockdowns tend to occur before any thorough discussion about how AI could actually enhance student creativity, comprehension, and inclusivity.

I haven't even touched on generative-AI imaging, video, music, and more. I have a little project in the works in this area so I will be returning to the subject – watch this space.

In my previous post – Farewell, the French Bulldog I looked at research from Oxford and Cambridge showing how LLMs, built on hoovering up the internet, may be staring down the barrel of model collapse if it is to keep feeding on what is rapidly becoming essentially its own outputs. This generation of LLMs may turn out to be as good as it gets.

Listening to journalists it seems they lie awake all night, bedsheets soaked in sweat, in full-blown panic that they, as well as creative writers and eventually the rest of us will be automated out of existence. Meanwhile, a certain brand of entrepreneur and C-suite execs are having wet dreams ot the prospect of robots doing all the work without needing to take time off or be paid. The likes of Elon Musk (if we can trust his judgment these days)¹⁹ and Sam Altman²⁰ think so. There is a certain fetishization in American business culture epitomized in the backslapping chortles of Musk and Donald Trump in a pre-election X-cast cajoling over how many people they had laid off, like two gym buddies boasting how much iron they lifted. Well that proved prophetic. But fear not, Silicon Valley has the answer: a universal basic income (UBI) of $1,000 per month should address the existential problem of humans, who will as a category become surplus to requirements in this fantasy economy²¹.

We are losing our minds over this thing. But nothing I have seen so far convinces me we are on the verge of being replaced.

Whether you are a journalist, writer, business analysts, software developers, lawyer, any profession you care to think of, the best thing you can do to get ahead in this new world is to think about how generative AI might be an accelerator in your day-to-day work. Even if you decide it is not for you, at least you will know what you are turning down. And you will know what you're up against, as others will certainly be using them.

We are in the middle of a tech revolution that is changing society in unrecognizable ways. We can look nervously to history for guidance at the impact of the printing presss or the industrial revolution.

But we have seen change as monumental as AI in our lifetimes - and we handled it.

If this is another step-change, we will all catch up sooner or later. Maybe you'll be an early adopter and get ahead of the curve, or maybe you will be like my dad, belatedly, complaining loudly every step of the way. Or maybe I am wrong, and this is just the latest hype surrounding an elaborate parlour trick. Make your own mind up.

One thing that is changing very rapidly though is just how convincing phishing and scame are becoming, as bad actors put this new tech to work.

It’s no longer enough to look for spelling mistakes or dodgy logos — some of these scams are already incredibly polished and personalized. It is not just ourselves, we need to look out for out parents and those less savvy, many of whom will be taken in by them. Is your voice out there on social media? Watch out for deep fakes that sound like you or a loved one²².

Whatever you make of it all, I think the best we we can do in the 21st century is to stay curious, keep learning, and remain vigilant against those who will always try to take advantage.