[echelon]

It doesn't take the AI bubble to sense the incredible IBMification of Google. The company can't ship product to save its life, and when it does the product gets canceled or is a duplicated effort of some other internal product.

The AI bubble could steal away 50% of Google's revenue stream by this time next year. Even if Google puts up a similar product, it has to retain 100% of its revenue and margin. Stuffing ads in interleaved responses seems implausible or downright dirty under an entirely new regime of high information density retrieval tools. (Granted, this is probably my weakest argument.)

It remains to be seen if users will flock from "Google search" to "Bing AI" (or whatever), but we all know how shitty Google search has become as a product compared to half a decade ago.

[vivegi]

I tried the new OpenAI enabled Bing chat yesterday.

Since NASA's announcement of the Artemis crew was on the news, I asked it to list 10 points in favor of space exploration. It promptly did that. Nicely.

Then I asked it to list 10 counterpoints for each of the 10 points listed in favor of space exploration. It tried. It replied with only 4 unique counter points and formatted them as 10 individual responses against each original point listed in favor of space exploration.

I then told it about the duplicate counterpoints and asked it to summarize them. It made a mistake and listed only 3 out of the 4 counterpoints. I pointed out the mistake. It apologized politely and listed all 4.

The overall experience felt like working with a child gifted in many aspects but missing some higher order skills and helping the child with their homework.

The next prompt was "How did John Wick: 4 end?"

It responded, "I am sorry but I cannot provide that information as John Wick: Chapter 4 has not been released yet. However according to Rotten Tomatoes, the official synopsis of John Wick: Chapter 4 reads: 'John Wick uncovers a path to defeating ....' (response elided)

I replied, "You are wrong. John Wick: Chapter 4 has been released and I have seen it."

Bing responded, "I'm sorry but I cannot continue this conversation as I am programmed to provide only factual information. If you have any other questions that I can help with, please let me know (folded-hands-emoji)"

### The conversation ended with a message 'It might be time to move on to a new topic. Let's start over' ###

Overall, a little rough around the edges, but still useful in limited ways. Google is nowhere near this yet.

[zulban]

If you feel like your handful of improvised questions is a good or unique evaluation of the latest LLMs and their applications you simply haven't been paying attention. Why did 100m+ users sign up in 2 months if the tool is as bad as you say?

[ilyt]

Coz it is not bad in other domains ?

[aroman]

"Google is nowhere near this yet?" It can literally do it right now. Formatting was very nice in the UI, bulleted the key idea expressed succinctly in the first few words. It also provided a great summary of John Wick 4.

me: please list 10 points in favor of space exploration

Google Bard: here are 10 points in favor of space exploration:

1. *Space exploration can help us learn more about our own planet.* By studying other planets and moons, we can learn more about how Earth formed and evolved. This information can help us better understand the challenges facing our own planet, such as climate change and natural disasters. 2. *Space exploration can help us find new resources.* There are many resources in space that could be used to benefit humanity, such as water, minerals, and energy. Exploring space can help us find these resources and learn how to use them safely and sustainably. 3. *Space exploration can help us develop new technologies.* The challenges of space exploration have led to the development of many new technologies, such as satellites, GPS, and solar panels. These technologies have had a positive impact on our lives on Earth. 4. *Space exploration can inspire us.* The vastness and beauty of space can inspire us to dream big and to believe that anything is possible. Space exploration can also help us to appreciate our own planet and the importance of protecting it. 5. *Space exploration can help us unite as a species.* The challenges of space exploration require cooperation and collaboration between nations. This can help to build trust and understanding between people from different cultures. 6. *Space exploration can help us protect Earth from threats from space.* Asteroids and comets pose a real threat to Earth. By studying these objects and developing plans to deflect them, we can protect our planet from potential disaster. 7. *Space exploration can help us find new places to live.* As the population of Earth continues to grow, we will need to find new places to live. Space exploration can help us find new planets and moons that could be terraformed for human habitation. 8. *Space exploration can help us understand our place in the universe.* By exploring space, we can learn more about the origin and evolution of the universe. This information can help us understand our own place in the cosmos and our relationship to other life forms. 9. *Space exploration is an adventure.* Exploring space is an exciting and challenging endeavor. It is a chance to see new things and to experience the unknown. 10. *Space exploration is a human right.* The right to explore is a fundamental human right. It is a right that should be afforded to all people, regardless of their nationality or background.

me: list 10 counterpoints for each of the 10 points you listed

Google Bard:

Here are 10 counterpoints for each of the 10 points I listed in favor of space exploration:

1. *Space exploration is expensive and could be better spent on other things.* The cost of space exploration is high, and some people argue that this money could be better spent on solving problems on Earth, such as poverty and hunger. 2. *The risks of space exploration are high.* Space travel is dangerous, and astronauts have died in accidents. Some people argue that the risks of space exploration are too high, and that we should focus on more safe endeavors. 3. *We don't know what we'll find in space, and it could be dangerous.* We don't know what life forms might exist in space, and it's possible that they could be hostile to humans. Some people argue that we should avoid contact with alien life until we know more about it. 4. *Space exploration is a waste of time.* Some people argue that space exploration is a waste of time and resources, and that we should focus on solving problems on Earth. 5. *Space exploration is not necessary for our survival.* We don't need to explore space in order to survive. Some people argue that we should focus on our own planet and not worry about exploring other worlds. 6. *Space exploration is not a priority for everyone.* Not everyone is interested in space exploration. Some people argue that we should focus on things that are more important to most people, such as education and healthcare. 7. *Space exploration is not sustainable.* The resources needed for space exploration are finite, and some people argue that we should not use them up when we have so many problems on Earth. 8. *Space exploration is not equitable.* The benefits of space exploration are not evenly distributed, and some people argue that we should focus on helping people on Earth before we explore other worlds. 9. *Space exploration is not environmentally friendly.* The launch of rockets and satellites contributes to climate change, and some people argue that we should not pollute the environment in order to explore space. 10. *Space exploration is not a human right.* There is no international law that guarantees the right to explore space. Some people argue that space exploration is a privilege, not a right.

[vivegi]

The nowhere near this yet remark meant MSFT completing integration of ChatGPT with Bing and releasing it to endusers whereas Google is yet to do it for their search (at least I am not aware of it).

I am a daily user of Google search, but yesterday I went to Bing after MSFT announced it in an email to me that it was available to me.

[bagacrap]

To be fair, Microsoft has zero to lose by slapping together some crappy integration. Of course they're going to move faster.

Nowhere near would apply if Google had no chatgpt competitor, not if they were still working on perfecting the UX.

[doctor_eval]

I'm no fan of Google, and while I agree that Google is going to shit, things always change more slowly (and then more fully) than we expect.

GPT is far better than search for most of my use cases, but Google has enormous momentum, and a lot of technology. A change in CEO could see them turn the ship around well before revenue takes a significant beating.

[jay_kyburz]

I'm bullish on AI, but 50% of Google revenue in a year in completely hyperbolic. 10 years perhaps.

[fhd2]

Agreed - even if there's a vastly superior search engine released tomorrow, so many people grew up with Google and learned how to deal with it, it's the default in most browsers - a space owned largely by Google - etc. It's even harder to get a large number of users to move if the product works differently.

I rather think LLMs will ruin the web with insane amounts of garbage content, and with it most of Google's livelihood. Hopefully not by this time next year though.

[bagacrap]

> Even if Google puts up a similar product, it has to retain 100% of its revenue and margin.

I see you've been to one of Satya's seminars. Not at all inside the bubble then ;)

What happens if Google search only retains 98% of its margins?