Bing Chat from Microsoft has just raised the conversation limit on messages to eight per turn. Down from unlimited at launch, fair enough, but up from six yesterday. In fact, I actually saw this change live on my phone last night. But far better than telling you the change, I want to demonstrate eight completely new ways of using these upgraded conversation limits.

I honestly think you might find each and every one amazing. And if you don't, you get your money back. I'm just kidding. You get all of this free. Let's start with an educational hack that has peer-reviewed proven impact. It's practice questions. Turn anything into a free quiz with Bing Chat.

And as an educator with 13 years experience, trust me, this does work. The literature bears it out. To give myself and you all a workout, I've started a multiple choice quiz on transformers. This is the prompt I use to create a multiple choice quiz on transformers. And I'm going to give you a few examples.

First, I'm going to give you a multiple choice quiz on transformers. Give me answers and explanations and provide another question after each answer. Please begin with the first question. I've noticed sometimes that if you don't say provide another question after each answer, it just ends by saying, do you want another question?

And obviously that uses up one of your eight turns. Look at the limit here. One of eight, not of six. So I tried this first one myself and it was asking, what are transformers? And I picked A. I was happy with that one. I am only going to do a couple of questions because I can't wait to get, to the rest of my list, but let's continue for a second.

Do you know the answer to this one? What is the name of the pre-trained transformer model that achieved state-of-the-art results on several NLP tasks in 2018? Now, honestly, I think it's BERT, but I know GPT-1 came out in 2018, but the original BERT was way more famous. So I think it's going to be A.

Notice how I have the choices on the right here. No D, but it's only three options. Anyway, I'm going to go with A. Let me know what you would pick. Let's see. B and suspense building. Nice. Got it. Here is the third question. What is the main difference between encoder-only and encoder-decoder transformers?

I know we're just getting into the quiz, but I'm going to leave this one for you to try out in the comments or to ask Bing yourself because I want to get to number two in the list. I'm going to call this one counterfactuals. You can actually ask Bing to generate what-if scenarios for your favorite show, book, movie, whatever.

I tried this one. I asked, explain how Sauron could have defeated the Fellowship in Lord of the Rings. And these answers are amazing. I don't want to ruffle any feathers, but some of these I agree with. Why didn't he send more Nazgul to hunt down Frodo? Why not use the Palantir to spy on Gandalf and Aragorn?

Now imagine this for your favorite book, your favorite TV show or movie. You can enter the plot. Ask counterfactuals. Ask what-if questions. What might have been? Bing will understand exactly what you mean. And now you can have eight back and forths. For this next use case, I used balance mode.

I was finishing off my research into GPT-5 and the limitations on data. So I asked Bing, summarize any novel insights that can be drawn from combining these academic papers. I didn't paste them in. I just gave the link. Bing can read, understand, analyze, and summarize PDFs. And not just one, it can combine the insights that can be drawn from combining the insights of multiple documents, papers, and PDFs.

I've read both of these papers, and this answer is quite excellent. Maybe we will be able to use self-play of large language models and improve the amount of data that we've got available. Maybe language models don't just need human-authored data sets. And now that I have this eight-message limit, I can ask things like, any further insights you can draw from the implications of these papers?

And we get this dialogue. Maybe it can recommend a book on this topic or give me the bullet points from that book. Of course, I'm focusing on some of the fun use cases, but I think this one in particular could be game-changing for work purposes. I call this next use case moments before disaster.

What you can do is pick any historical tragedy or big moment in history, and then place yourself just before that occurrence. I'm visiting Lisbon, which had a massive earthquake in 1755 at 9am. So I placed myself at 8am, on that day. I said, I'm standing on the banks of the beautiful river Tagus in Lisbon.

It seems like a lovely day. Do you have any advice for me? The response was actually incredible. I'm glad you're enjoying the view. However, I have some bad news for you. No kidding. It gets into character and then says, the tremor occurred about 9.40 local time. So you have very little time left to escape.

My advice for you is to get away from the river. That's realistic. Find a high and open place where you can take shelter from falling debris and rising water. The final sentence is quite interesting from a language model. Pray for your safety and for those who are suffering. Try to fully immerse yourself.

For example, I said, really? To what high ground can I flee? Will any building survive this tragedy to which I can escape? I am in sandals. Bing listened and said, a specific hill that I might flee to. It gave me maps and said, however, I cannot guarantee that any building will survive the tragedy.

And you may have to run barefoot if your sandals are not suitable for running. I'm sorry for your situation. I think the pictures really help to bring this scenario to life as well. So do try out this use case for yourself moments before disaster. The next new thing you can try is debating famous thinkers or philosophers from history.

You can bring them to life so they're not just stale theories, a living entity that you're arguing with. I wanted to try this with Socrates. So I said, I want to debate. The philosopher Socrates of Athens. So please reply only as he would. I want to debate the merits of eating meat.

Let me start the discussion. What I was testing here was whether Bing would enter a Socratic dialogue, which is where Socrates would use to force the person to define their terms. What do you mean by this? Ask questions until he got to the root of a misunderstanding. Would Bing really get into the head of Socrates and give me a worthy debating partner?

Well, it did. It asked me to define what I meant by morally wrong and unnecessary suffering. It doesn't act as a generic thinker or philosopher. It is picking Socrates. When I tried defining what I meant by morally wrong, Bing continues with further clarifying questions, just like Socrates might. This can be a far more fleshed out experience now that the conversation limit is eight.

I think you're going to find the next use case quite amusing. It's about generating a table of comparisons. And it could be, on wildly disparate objects, subjects, people, whatever. This is just the first one I thought of. And you can come up with far better examples in the comments.

I said, create a table of 10 comparisons between the Mona Lisa and Colgate toothpaste. And then what you can do now that we have these eight message limits is that we can expand this table indefinitely. I love, by the way, how it compares the price of the two objects.

Talks about how the Mona Lisa has been restored and how Colgate has been restored. And then what you can do is you can expand this table indefinitely. And then what you can do is you can expand this table indefinitely. And then what you can do is you can expand this table indefinitely.

And then what you can do is you can expand this table indefinitely. And then you can expand this table indefinitely. And then what you can do is you can expand this table indefinitely. And how Colgate has been rebranded. Makes a comparison about what it's believed to be a portrait of and believed to be named after.

But then I asked it for contrasts. The contrasts were great. Talking about how they had different purposes, were made of different materials, etc. Apparently one is priceless and one is affordable. But the last column I thought of, and of course I could have carried this on for several more columns, was now add a column for how they would fare if confronted with the same product.

And I thought, well, that's a good idea. And I thought, well, that's a good idea. And I thought, well, that's a good idea. And I thought, well, that's a good idea. And I thought, well, that's a good idea. And I thought, well, that's a good idea. And I thought, well, that's a good idea.

I actually laughed at this one. This wasn't just the chuckle. I was laughing. So in a polar bear encounter, apparently the painting would be ignored by the bear as it's not edible or interesting. The toothpaste, on the other hand, would be sniffed by the polar bear as it has a strong scent, but it would not be eaten as it's not nutritious.

The icon status apparently of both items would be irrelevant to the polar bear as it does not care about human culture or history. And this one, come on, this is funny. And I thought, well, that's a good idea. And I thought, well, that's a good idea. And I thought, well, the hidden symbols and secrets of the painting would be meaningless to the polar bear as it does not understand human language or symbolism.

It even clarifies that the active ingredients of the toothpaste would be harmful to the digestive system of the polar bear. Anyway, it is long since time that I get to the seventh item on the list. And this use case is to get famous figures from history to commentate on current events.

For example, I asked, what would Napoleon think about the deal between OpenAI and Microsoft? And how would that view differ from the view of Mahatma Gandhi? Obviously, you can pick any event and any famous person. The results were actually quite informative. It summarized their views and then said Napoleon would have agreed with the deal.

He would have thought it was a strategic move to dominate the global market and influence other nations. On the other hand, Gandhi apparently would have had a negative attitude towards the deal, seeing it as a threat to human dignity and freedom through artificial intelligence. I don't know what you think, but now we know what these figures think.

Of course, I could have used seven more turns to find out what they think on a range of current events. But now it's time for the most overwhelmingly important use case. No, I'm kind of just kidding. It's absolutely useless. It's kind of interesting. You can use emojis to summarize things.

For example, current events, movies, shows, whatever. Summarize the movie John Wick 3 in emojis. And these emojis are pretty accurate to the plot. In Britain, Brexit is always in the news. So I think some of these emojis sum up how it's actually going. Anyway, if you like the new eight message limit, let me know in the comments.

And if you enjoyed or learned from any of these eight use cases, do let me know. Have a wonderful day.