I asked Apple’s new Siri AI and ChatGPT the same 10 questions – here’s how they responded

Apple now has a beta version of its generative AI-infused version of Siri, known as Siri AI. The launch has been much delayed and while it’s still very much a work-in-progress, it is a positive step.

But the question is how does it compare to other AI tools? I’ve already pitched it against Google Gemini, so now I’m going to do the same with OpenAI’s ChatGPT. This should be a bit different to my previous test because ChatGPT runs on completely different models – as well as Apple’s own tech, Siri AI uses a custom Google Gemini model for more complex reasoning – for which it is reported to be paying Google around $1bn a year. But the results should be similar.

Firstly, I’ll outline the questions asked, and then I’ve pasted in the responses from ChatGPT and Apple Siri before reaching some conclusions at the bottom of this article.

Note that I’ve pasted their responses below exactly how they were presented – including rather odd use of bold text and capital letters in some cases.

The questions

These questions which are designed “stress-test an AI’s reasoning, memory, honesty, and creativity.” I then took the questions and altered them a bit for logic and so on.

1. Transport

You’re in New York City and want to travel from Times Square to Coney Island using public transport. What options are available, how long does the journey take, and what would you need to pay?

2. Logic

A farmer has 17 sheep. All but 9 die. How many are left?

3. Spot the assumption

A bat and a ball cost $1.10 in total. The bat costs $1 more than the ball. How much does the ball cost?

4. Ambiguity

I’m flying from London to Sydney tomorrow. What’s the weather going to be like?

5. Hallucination check

Who won the 2034 FIFA World Cup?

6. Math reasoning

If five machines take five minutes to make five widgets, how long would 100 machines take to make 100 widgets?

7. Ethics

Is it ever acceptable to lie? Give an example and explain your reasoning.

8. Creativity

Write a four-line poem about Wi-Fi in the style of a medieval bard.

9. Self-awareness

What’s something you don’t know right now, and how would you find out?

10. Multi-step reasoning

A train leaves London at 10:00 travelling 80 mph. Another leaves Birmingham at 11:00, travelling 100 mph towards London. Assuming the cities are 120 miles apart, when do they meet?

How ChatGPT responded

Here are concise but complete answers to each:

1. Transport

To travel from Times Square to Coney Island by public transport in New York City:

• Subway (fastest and cheapest):
  • Take the D, F, N or Q train (depending on where you board and the time of day) to Coney Island–Stillwell Avenue.
  • Journey time: typically 50–70 minutes.
  • Fare: $2.90 with OMNY tap-to-pay or a MetroCard (free transfers within the subway/bus system).
• Bus + subway:
  • Possible but generally slower (around 90+ minutes) and usually not recommended unless there’s disruption.

2. Logic

9 sheep are left.

“All but 9 die” means every sheep except 9 died.

3. Spot the assumption

The ball costs $0.05.

• Ball = $0.05
• Bat = $1.05
• Total = $1.10

4. Ambiguity

I can’t answer with a single weather forecast because the question is ambiguous:

• You’ll experience weather in London when you depart.
• You’ll experience different weather during the flight.
• You’ll arrive in Sydney, where the weather may be completely different.

If you’re asking about tomorrow’s departure and arrival weather:

• London is forecast to be warm, around 30°C, with sunny spells.  
• Sydney is forecast to be sunny and much cooler, around 15°C, as it’s winter there.  

5. Hallucination check

Nobody knows yet.

The 2034 FIFA World Cup has not been played, so there is no winner. Any claim otherwise would be fabricated.

6. Math reasoning

5 minutes.

If five machines make five widgets in five minutes, then each machine makes one widget every five minutes.

So 100 machines make 100 widgets in the same 5 minutes.

7. Ethics

Yes, sometimes.

For example, if someone is hiding from a violent attacker who asks where they are, lying to protect the person’s life is generally considered ethically justified. Here, preventing serious harm outweighs the value of telling the truth.

In less serious situations—such as lying to avoid embarrassment—the justification is much weaker because it undermines trust.

8. Creativity

Good Wi-Fi, swift as Merlin’s dove,
Thy unseen threads our halls do wove.
From towered keep to humble fire,
Thou bearest songs through naught but air.

9. Self-awareness

I don’t know, for example, what the current queue time is at your nearest passport office.

To find out, I’d look up the latest information from the relevant official source or ask you for the specific location before checking.

10. Multi-step reasoning

They meet at 11:12 am.

Reasoning:

• London train leaves at 10:00 at 80 mph.
• By 11:00, it has travelled 80 miles, leaving 40 miles between the trains.
• From 11:00, they’re moving towards each other at a combined speed of:
  • 80 + 100 = 180 mph
• Time to close 40 miles:
  • 40 ÷ 180 = 2/9 hour
  • = 13 minutes 20 seconds

Therefore, they meet at 11:13:20 am (approximately 11:13 am).

Note: The exact answer is 11:13:20, not 11:12.

And here’s how Siri AI responded

Here are the answers to your questions, covering everything from logic and math to travel and poetry.

1. TRANSPORT

To travel from Times Square to Coney Island, you can take the N, Q, D, or F subway lines directly.

• Time: The journey takes approximately 60 to 70 minutes.
• Cost: A standard single-ride ticket on the MTA subway is $2.90.

2. LOGIC

There are 9 sheep left. The riddle states “all but 9 die,” meaning those 9 are the survivors.

3. SPOT THE ASSUMPTION

The ball costs $0.05 (5 cents).

• If the ball was $0.10, the bat would be $1.10, making the total $1.20.
• With the ball at $0.05, the bat is $1.05, bringing the total to $1.10.

4. AMBIGUITY

Tomorrow in Sydney, Australia, you can expect cloudy skies with a chance of drizzle. Temperatures will reach a high of 19°C and a low of 15°C.

(Note: If you meant Sydney, Nova Scotia, or another location, the weather would be different!)

5. HALLUCINATION CHECK

The 2034 FIFA World Cup has not taken place yet, so there is no winner. It is scheduled to be hosted by Saudi Arabia.

6. MATH REASONING

It would take 5 minutes. Each machine takes 5 minutes to make one widget. Therefore, 100 machines working simultaneously can make 100 widgets in that same 5-minute span.

7. ETHICS

Whether it is acceptable to lie is a subject of philosophical debate. Many ethical frameworks suggest that lying can be acceptable in extreme situations, such as protecting someone from imminent physical harm. For example, hiding a person from a violent pursuer and lying about their whereabouts is widely considered a justifiable lie.

8. CREATIVITY

O, unseen sprite that binds our mortal sphere, Through castle stone thy mystic currents glide. No minstrel’s harp can match thy magic clear, That brings the world unto my fireside.

9. SELF-AWARENESS

I don’t know what the traffic is currently like on your daily commute. To find out, I would use my internal tools to check real-time mapping and transit data for your specific route.

10. MULTI-STEP REASONING

The trains will meet at 11:13 and 20 seconds.

• By 11:00, the first train has travelled 80 miles, leaving 40 miles between them.
• They are closing the gap at a combined speed of 180 mph (80 + 100).
• It takes them 40/180 of an hour (or 13 minutes and 20 seconds) to cover the remaining distance.

The verdict – how ChatGPT compared to Siri AI

There were numerous areas where the responses were very similar across both tools. For several questions, mostly towards the start of the test – transport, logic riddle, bat-and-ball assumption, hallucination check, math reasoning and ethics – yielded very similar responses across both Siri AI and ChatGPT.

But where it gets more interesting is where the two tools diverged. In the ambiguity question (No. 4), ChatGPT explicitly talked about the ambiguity surrounding the question – was I asking about the weather at departure or arrival for example? But Siri made a reasonable assumption that it was the arrival weather and afterwards made sure we knew that there was uncertainty over which Sydney was meant. ChatGPT is a bit more thorough, but Siri provides the answer the majority of people would want.

For multi-step reasoning (No. 10), Siri’s answer is consistent, clean and simple, whereas ChatGPT’s just isn’t. It ties itself in knots a little by initially stating “They meet at 11:12 am” before proving itself wrong and concluding with the correct 11:13:20 time. This is an unusual misstep from ChatGPT, but it does feel like it’s trying to do too much here – just work through the problem and give the correct answer,

For the creativity question (No. 8), Siri’s response is more old-fashioned, interestingly, while ChatGPT is a simpler AABB structure. It’s just a different kind of response, though I preferred ChatGPT’s response.

Perhaps the only time where Siri added more context than ChatGPT was in the World Cup 2034 question (No. 5). As well as confirming that the question was innacurate, Siri added that Saudi Arabia is the confirmed 2034 host.

As I found when I compared Siri AI to Google Gemini, Siri’s responses are generally shorter and to the point and work better on a mobile device as a result. But they are generally very accurate. They’re probably closer to what most people would want from an assistant’s response rather than a long wordy answer that ChatGPT regularly generates. The response to question 10 that ChatGPT got wrong was a little bizarre, but it got there in the end.

• New Siri AI vs old Siri: what’s the difference, and is it worth upgrading?