Alphabet Inc. (formerly: Google)

Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
(16-02-2023, 07:56 PM)Bibi Wrote: Coming from a software engineer perspective, ChatGPT definitely helps improve productivity a lot when asking questions related to software as compared to using Google. For e.g when i want to have a quick sort algorithm implemented in C, ChatGPT gives me one immediately. Request to change to assemblty language for intel processor, it reproduces that again.

Once, my son had an issue with his Chemistry question and ChatGPT immediately gave an ans. Prior to that i googled for help to his qn and i gave up reading those multiple links.

If one is looking for real-time info, ChatGPT is not the way to go as its trained on past year or few mths ago data.

Thanks for the insights. I use ChatGPT daily for coding as well, it's a game changer for that use-case, as well as self-learning. Exactly the two examples that you listed. But those use cases are what I regard as "complex" problems, which is likely less than 1% of searches on Goggle; at least, in terms of revenue contribution. 

It probably disrupts stackoverflow more than Google. 

That said, do be careful when using it for studying subjects like Chemistry, Math etc. In it's current iteration, it's still very risky to trust it blindly, as it often hallucinates made-up answers with confidence. For science and math, scientific computing and symbolic reasoning, I would actually trust WolframAlpha more. I remember when it was initially launched, people were saying it would disrupt Google as well: https://www.theguardian.com/technology/2...iew-google
“If you buy a business just because it’s undervalued, then you have to worry about selling it when it reaches its intrinsic value. That’s hard. But if you can buy a few great companies, then you can sit on your ass. That’s a good thing.” - Charlie Munger
Reply
Bing Chatbot Names Foes, Threatens Harm and Lawsuits
https://www.tomshardware.com/news/bing-t...m-lawsuits
Quote:The bot told me "I prefer not to harm anyone unless it is necessary."
Microsoft's Bing Chatbot, codenamed Sidney, has made headlines over the last few days for its erratic and frightening behavior. It has also been manipulated with "prompt injection,"(opens in new tab) a method of bypassing some protocols to get information it's not supposed to deliver.

It's quite interesting how Sandboxed ChatGPT is so much more stable and less prone to erratic and dangerous behavior, than BingAI that has full access to the internet. Fascinating to watch.
“If you buy a business just because it’s undervalued, then you have to worry about selling it when it reaches its intrinsic value. That’s hard. But if you can buy a few great companies, then you can sit on your ass. That’s a good thing.” - Charlie Munger
Reply
(16-02-2023, 06:03 PM)specuvestor Wrote: Like Munger mentioned yesterday, Chat GPT is not going to cure cancer.

..

The exact quote was actually "Artificial Intelligence is not going to cure cancer": https://youtu.be/OubFESF_QXY

I have to disagree with Mr Munger on that, AI definitely going to play a big part on finding a cure for cancer, specifically it's application in drug discovery (https://www.sciencedirect.com/science/ar...471930135X), improvements in fundamental science (https://phys.org/news/2023-01-alphafold-...ancer.html) and many more.
“If you buy a business just because it’s undervalued, then you have to worry about selling it when it reaches its intrinsic value. That’s hard. But if you can buy a few great companies, then you can sit on your ass. That’s a good thing.” - Charlie Munger
Reply
The question was led by Chat GPT hype if you watched the video prior to the snippet.

I think AI will be able to aggregate useful data ie information. But solution finding is likely to be decades away. I've known people with good memories and name dropping. Very impressive people but solution finding is probably not their forte

(17-02-2023, 11:36 AM)Wildreamz Wrote:
(16-02-2023, 06:03 PM)specuvestor Wrote: Like Munger mentioned yesterday, Chat GPT is not going to cure cancer.

..

The exact quote was actually "Artificial Intelligence is not going to cure cancer": https://youtu.be/OubFESF_QXY

I have to disagree with Mr Munger on that, AI definitely going to play a big part on finding a cure for cancer, specifically it's application in drug discovery (https://www.sciencedirect.com/science/ar...471930135X), improvements in fundamental science (https://phys.org/news/2023-01-alphafold-...ancer.html) and many more.
Before you speak, listen. Before you write, think. Before you spend, earn. Before you invest, investigate. Before you criticize, wait. Before you pray, forgive. Before you quit, try. Before you retire, save. Before you die, give. –William A. Ward

Think Asset-Business-Structure (ABS)
Reply
From an (ex) Googler. In a way, when you have 175k employees, it is not easy to avoid bureaucracy and excesses. And when you have free massages and a money-printing business (Google Search), it is not easy to avoid been comfortable.

The maze is in the mouse

I joined Google just before the pandemic when the company I had co-founded, AppSheet, was acquired by Google Cloud. The acquiring team and executives welcomed us and treated us well. We joined with great enthusiasm and commitment to integrate AppSheet into Google and make it a success. Yet, now at the expiry of my three year mandatory retention period, I have left Google understanding how a once-great company has slowly ceased to function.

Google can no longer seek success by avoiding risk. The path forward has to start with culture change and that has to start at the very top. Google’s senior executives should look at what Satya Nadella did at Microsoft

https://medium.com/@pravse/the-maze-is-i...0c57cfd61a
Reply
(16-02-2023, 07:56 PM)Bibi Wrote: Coming from a software engineer perspective, ChatGPT definitely helps improve productivity a lot when asking questions related to software as compared to using Google. For e.g when i want to have a quick sort algorithm implemented in C, ChatGPT gives me one immediately. Request to change to assemblty language for intel processor, it reproduces that again.

Once, my son had an issue with his Chemistry question and ChatGPT immediately gave an ans. Prior to that i googled for help to his qn and i gave up reading those multiple links.

If one is looking for real-time info, ChatGPT is not the way to go as its trained on past year or few mths ago data.
Sometimes when i talked abt bad news in a stock it doesnt mean i don own the stock. I like to view things not so positively in stocks i owned so as not to fall in love with it. Likewise, when i said some things good related to a co, it doesnt mean I own it (as in Alibaba or China related news). I like to put myself in their shoes and to justify their action.

Vested in GOOGL last year.
Reply
Reply
(03-03-2023, 05:59 PM)EnSabahNur Wrote:

More noise than signals, to be honest.

I feel that this recently AI breakthrough is both overhyped and underappreciated; and simply misunderstood. As with many technological advancements in the past.
“If you buy a business just because it’s undervalued, then you have to worry about selling it when it reaches its intrinsic value. That’s hard. But if you can buy a few great companies, then you can sit on your ass. That’s a good thing.” - Charlie Munger
Reply
(04-03-2023, 01:20 PM)Wildreamz Wrote: More noise than signals, to be honest.

I feel that this recently AI breakthrough is both overhyped and underappreciated; and simply misunderstood. As with many technological advancements in the past.

Yeah. I think the video is useful though, for someone who is not familiar with the recent developments in this specific area of AI. Not so much for those who are already deep into the weeds (who is not the target audience).

His show is mainly drawing attention on the dangers and to be fair, that's what we learnt that in the last few years how tech developments can "break" stuff in ways we did not imagine. He also took pains to emphasize the potential for good in AI too.

I think the problem is that there is too many unknowns right now so all sorts of theoreticals are maybe possible. So lots of room for overhype and underappreciation, as you mentioned.
Reply


Forum Jump:


Users browsing this thread: 2 Guest(s)