Politics & Stuff

[Fulham Broadway 17,306]

4 hours ago, cosmicway said:

Rothmans football book -

haha I remember those !

[Sir Mikel OBE 4,920]

20 minutes ago, MoroccanBlue said:

AI will not enter data/technology industries for a very, very long time. If ever. 

Its already there realistically. The models we use to train AI to find anomalies in Security is already being rolled out by M365 users with security copilot.

Github's AI copilot also generated about 46% of developer's code as of earlier this year:https://github.blog/2023-02-14-github-copilot-now-has-a-better-ai-model-and-new-capabilities/

The question is will it take ALL the jobs, which it wont, but it will take enough to put a dent in the industry and that dent is already showing in technology in America.

[MoroccanBlue 5,378]

8 minutes ago, Sir Mikel OBE said:

Its already there realistically. The models we use to train AI to find anomalies in Security is already being rolled out by M365 users with security copilot.

Github's AI copilot also generated about 46% of developer's code as of earlier this year:https://github.blog/2023-02-14-github-copilot-now-has-a-better-ai-model-and-new-capabilities/

The question is will it take ALL the jobs, which it wont, but it will take enough to put a dent in the industry and that dent is already showing in technology in America.

What I meant was organizations in these industries will never allow AI to directly pull or access data. They are also unreliable when it comes to interpreting and analyzing data sets. 

Data and IS careers are safe. 

Edited August 11, 2023 by MoroccanBlue

[Sir Mikel OBE 4,920]

2 hours ago, MoroccanBlue said:

What I meant was organizations in these industries will never allow AI to directly pull or access data. They are also unreliable when it comes to interpreting and analyzing data sets. 

Data and IS careers are safe. 

I think you are speaking more from hope than expectations. Security copilot that Microsoft uses to manage vulnerabilities worldwide(including MENA in Israel for example with Sentinel) is directly using security and data in their libraries to recognize potential cybersecurity threats. All of this is maintained in data sets accessible to their language models.

There will always be someone required to handle some data sets, but I already see jobs being replaced by these tools. Its lowering the amount of people needed for it.

[robsblubot 3,595]

Finally a chance to discuss something related to my realm of expertise (networking/internet backbone);
All jobs can be replaced by AI, and that will be the big negative impact AI will have short/med-term: never yet seen job loss (order of magnitude larger than of the indust. rev). All jobs are replaceable, but some are MUCH easier and imminent than others (commercial drivers will be hit hard).

AI "readyness' is IMHO exaggerated at this time. Detecting whether AI is working correctly will remain a challenge for quite some time.

There will be cases where the simple question would be, "do we want to have AI here?" and the answer may very well be no from social/humane reasons.

Watching the media talk about Terminator-like doomsday scenarios is honestly laughable, esp coming from "serious" folks.

Edited August 11, 2023 by robsblubot

[MoroccanBlue 5,378]

51 minutes ago, Sir Mikel OBE said:

I think you are speaking more from hope than expectations. Security copilot that Microsoft uses to manage vulnerabilities worldwide(including MENA in Israel for example with Sentinel) is directly using security and data in their libraries to recognize potential cybersecurity threats. All of this is maintained in data sets accessible to their language models.

There will always be someone required to handle some data sets, but I already see jobs being replaced by these tools. Its lowering the amount of people needed for it.

I'm speaking from experience. I work in Data. I've been an analyst and worked in data science contracts. Engineers, Analysts, scientists. These careers will always be protected. (unless you work in big tech atm)

Giving access to ChatGPT or any OpenAI tool to query or pull data won't ever be a reality if you work with a company that handles client information. Which is basically all of them if you need a data team. It's just not going to happen on an ethical standpoint. It's simply a data breach waiting to happen. 

They've also tested these OpenAI tools to analyze and interpret data sets, and they've been widely off the mark. 

[cosmicway 1,333]

8 hours ago, Sir Mikel OBE said:

The internet wont die, but it will change fundamentally.


The good thing is, we have lives through multiple iterations of the internet. I am old enough to remember how different it was even this century before social media took over.

What is more realistic is that AI will kill a bunch of jobs. Fields like technical writing, web development, quality assurance(code testers), copywriting, and even business consulting are in the firing zone of being replaced. This will fundamentally change the global economy. You are already seeing ripples of it in Big tech. Back when I graduated from College ~10 years ago you could literally apply for a software engineering job the week of graduation, and the next week walk into an entry level job that was paying 60k back when average household income was 50k. you were middle class from the start. Now? These poor kids are entering a saturated field and going up against machines that can easily take away jobs from many. AI programs very well so when we used to need 10 or so undergrads to work on a system we can hire 2 and use AI to make up the difference. Same story happening in design, writing, data entry...ect.

Prices are going up too.

 

Jobs die all the time.
I remember this scene: I go to the army recrruitment office -they called me rather- and I have to fill certain forms. So out in the street there is a row of small tables where the "application writers" were sitting. It was two rows 500 meters long each. You pay 5 drachmas to one of the old men to fill your forms and then you go upstairs.
The same thing outside the courthouses. In the courthouses it was a neighbourhood of typist offices who were also making photocopies.
Now all those have disappeared - it's vegan restaurants now. 
What made them disappear ? Microsoft word.

What about software engineering ? I wrote a certain app for Amstrad 128 then showed it to the Amstrad representative. As we were chatting I asked what manner of software they have for the Amstrad, other than mine and other than pacman-spaceshiips for three year olds. They had nothing really. The man said to me that he was writing some program for the electric installations of buildings - that exciting.
The computer age was just starting. After some time all manner of developers made their appearance. With a video club management program one could strike it rich.
But after windows xp all this died too. Because using excel one could manage his video collection with a little training - they did n't have to go to the market.
Well software engineering did n't die but it was no longer easy money.

The same with AI. We are like the men with the tea spoons who wanted to dig Trafalgar square.


 

[Vesper 30,119]

The Godfather in Conversation: Why Geoffrey Hinton is worried about the future of AI

Geoffrey Hinton, known to many as the “Godfather of AI,” recently made headlines around the world after leaving his job at Google to speak more freely about the risks posed by unchecked development of artificial intelligence, including popular tools like ChatGPT and Google’s PaLM. Why does he believe digital intelligence could hold an advantage over biological intelligence? How did he suddenly arrive at this conclusion after a lifetime of work in the field? Most importantly, what – if anything – can be done to safeguard the future of humanity? The University of Toronto University Professor Emeritus addresses these questions and more in The Godfather in Conversation.

 

Geoffrey Everest Hinton CC FRS FRSC^[12] (born 6 December 1947) is a British-Canadian cognitive psychologist and computer scientist, most noted for his work on artificial neural networks. From 2013 to 2023, he divided his time working for Google (Google Brain) and the University of Toronto, before publicly announcing his departure from Google in May 2023 citing concerns about the risks of artificial intelligence (AI) technology.^[13] In 2017, he co-founded and became the chief scientific advisor of the Vector Institute in Toronto.^[14][15]

With David Rumelhart and Ronald J. Williams, Hinton was co-author of a highly cited paper published in 1986 that popularised the backpropagation algorithm for training multi-layer neural networks,^[16] although they were not the first to propose the approach.^[17] Hinton is viewed as a leading figure in the deep learning community.^{[18][19][20][21][22]} The dramatic image-recognition milestone of the AlexNet designed in collaboration with his students Alex Krizhevsky^[23] and Ilya Sutskever for the ImageNet challenge 2012^[24] was a breakthrough in the field of computer vision.^[25]

Hinton received the 2018 Turing Award (often referred to as the "Nobel Prize of Computing"), together with Yoshua Bengio and Yann LeCun, for their work on deep learning.^[26] They are sometimes referred to as the "Godfathers of AI" and "Godfathers of Deep Learning",^[27][28] and have continued to give public talks together.^[29][30]

In May 2023, Hinton announced his resignation from Google to be able to "freely speak out about the risks of A.I."^[31] He has voiced concerns about deliberate misuse by malicious actors, technological unemployment, and existential risk from artificial general intelligence.^[32]

[Sir Mikel OBE 4,920]

16 hours ago, MoroccanBlue said:

I'm speaking from experience. I work in Data. I've been an analyst and worked in data science contracts. Engineers, Analysts, scientists. These careers will always be protected. (unless you work in big tech atm)

Giving access to ChatGPT or any OpenAI tool to query or pull data won't ever be a reality if you work with a company that handles client information. Which is basically all of them if you need a data team. It's just not going to happen on an ethical standpoint. It's simply a data breach waiting to happen. 

They've also tested these OpenAI tools to analyze and interpret data sets, and they've been widely off the mark. 

We are halfway here as Microsoft Enterprise copilot as its already guaranteed business data is secure: https://blogs.microsoft.com/blog/2023/03/16/introducing-microsoft-365-copilot-your-copilot-for-work/

I do agree OpenAI is still hit or miss with data(especially the new 4 models) but it is being tweaked constantly. The key is not making it perfect. The key is making it good enough to where we can replace a team of 5 data scientists with 2 or 1. That is where the job loss will happen, and where it is already happening in other fields.

The most sobering thing to happen to me this summer at least was seeing all of our interns at Microsoft work on these AI projects all happy, and then realize now that the internship is over where in the past years 80%+ of them had full time job offers for the next summer that next to none are getting offers now. We just dont have the need anymore.

[Sir Mikel OBE 4,920]

17 hours ago, robsblubot said:

Finally a chance to discuss something related to my realm of expertise (networking/internet backbone);
All jobs can be replaced by AI, and that will be the big negative impact AI will have short/med-term: never yet seen job loss (order of magnitude larger than of the indust. rev). All jobs are replaceable, but some are MUCH easier and imminent than others (commercial drivers will be hit hard).

AI "readyness' is IMHO exaggerated at this time. Detecting whether AI is working correctly will remain a challenge for quite some time.

There will be cases where the simple question would be, "do we want to have AI here?" and the answer may very well be no from social/humane reasons.

Watching the media talk about Terminator-like doomsday scenarios is honestly laughable, esp coming from "serious" folks.

We won't judge AI readiness on being perfect though. We judge it how we judge the rest of business. Can it be good enough for an MBA to show as a cost saving metric?

I think commercial driving is more safe than something like copywriting or web development because a truck going wrong on the road can still lead to immediate death. White collar work in general is infinitely in more danger at the current moment.

[Sir Mikel OBE 4,920]

11 hours ago, cosmicway said:

 

Jobs die all the time.
I remember this scene: I go to the army recrruitment office -they called me rather- and I have to fill certain forms. So out in the street there is a row of small tables where the "application writers" were sitting. It was two rows 500 meters long each. You pay 5 drachmas to one of the old men to fill your forms and then you go upstairs.
The same thing outside the courthouses. In the courthouses it was a neighbourhood of typist offices who were also making photocopies.
Now all those have disappeared - it's vegan restaurants now. 
What made them disappear ? Microsoft word.

What about software engineering ? I wrote a certain app for Amstrad 128 then showed it to the Amstrad representative. As we were chatting I asked what manner of software they have for the Amstrad, other than mine and other than pacman-spaceshiips for three year olds. They had nothing really. The man said to me that he was writing some program for the electric installations of buildings - that exciting.
The computer age was just starting. After some time all manner of developers made their appearance. With a video club management program one could strike it rich.
But after windows xp all this died too. Because using excel one could manage his video collection with a little training - they did n't have to go to the market.
Well software engineering did n't die but it was no longer easy money.

The same with AI. We are like the men with the tea spoons who wanted to dig Trafalgar square.


 

Yes sir.

 

Amstrad is super old school my man, but yea it was easy money in my day so I cant imagine how it was in the very old days. Are you retired now?

[MoroccanBlue 5,378]

1 hour ago, Sir Mikel OBE said:

We are halfway here as Microsoft Enterprise copilot as its already guaranteed business data is secure: https://blogs.microsoft.com/blog/2023/03/16/introducing-microsoft-365-copilot-your-copilot-for-work/

I do agree OpenAI is still hit or miss with data(especially the new 4 models) but it is being tweaked constantly. The key is not making it perfect. The key is making it good enough to where we can replace a team of 5 data scientists with 2 or 1. That is where the job loss will happen, and where it is already happening in other fields.

The most sobering thing to happen to me this summer at least was seeing all of our interns at Microsoft work on these AI projects all happy, and then realize now that the internship is over where in the past years 80%+ of them had full time job offers for the next summer that next to none are getting offers now. We just dont have the need anymore.

I’m highly skeptical in organizations allowing a third party to directly query and pull their data, nor do I see it even being listed as a feature for 365 Copilot. It’s simply a tool to analyze/automate already pulled data. Pulling and querying data is something that will always be a human element. Unless they replace their entire data platform for Microsoft Fabric. 
 

Much like Excel, I see Copilot making a lot of jobs easier vs making them redundant, where instead of waisting 30-40% of the time on needless tasks they can focus all of it on business driven needs. 

Edited August 12, 2023 by MoroccanBlue

[cosmicway 1,333]

39 minutes ago, Sir Mikel OBE said:

Yes sir.

 

Amstrad is super old school my man, but yea it was easy money in my day so I cant imagine how it was in the very old days. Are you retired now?

No, not really retired.
I bought my first machine in 1983. 
 

[Sir Mikel OBE 4,920]

1 hour ago, MoroccanBlue said:

I’m highly skeptical in organizations allowing a third party to directly query and pull their data, nor do I see it even being listed as a feature for 365 Copilot. It’s simply a tool to analyze/automate already pulled data. Pulling and querying data is something that will always be a human element. Unless they replace their entire data platform for Microsoft Fabric. 
 

Much like Excel, I see Copilot making a lot of jobs easier vs making them redundant, where instead of waisting 30-40% of the time on needless tasks they can focus all of it on business driven needs. 

And then it depends on how your businesses operate.

AI does make jobs easier, but that does not mean it frees the same workers to other things. It simply means we get rid of plenty of workers and place the remaining "less workload" on fewer and fewer people. That becomes the true cost of AI.

[Sir Mikel OBE 4,920]

1 hour ago, cosmicway said:

No, not really retired.
I bought my first machine in 1983. 
 

I cant imagine how the industry has changed in your time.

 

Honestly I dont see myself wanting to keep up with it over time. Im trying to do this for another 10 or so years and be done in my early 40s. My first machine was my fathers win 95 machine, in which I learned Visual Basic. He just wanted me to get a job that paid better than driving delivery trucks that he had. Never knew it would end up with me working for them lol.