96 replies to this topic

[todd]

Fear sells

That’s the scary part.

[Victoria Watcher]

Sam Altman will return as CEO of OpenAI, the startup said early Wednesday morning on X, formerly known as Twitter. The move follows immense pressure from employees and investors on the board that ousted him less than a week ago.

 

https://www.cnbc.com...ter-ouster.html

Edited by Victoria Watcher, 21 November 2023 - 10:26 PM.

[dkuitu]

 

Elon Musk launches AI chatbot 'Grok’ — says it can outperform ChatGPT

 

Grok costs $16 per month on X Premium Plus. But for now it is only offered to a limited number of users in the United States.

 

https://cointelegrap...perform-chatgpt

 

 

 

 

Musk and the xAI team said a “unique and fundamental advantage” possessed by Grok is that it has real-time knowledge of the world via the X platform.

 

“It will also answer spicy questions that are rejected by most other AI systems,” Muska and xAI said. "Grok is designed to answer questions with a bit of wit and has a rebellious streak, so please don’t use it if you hate humor!"

 

 

 

He had a grok demo on one of his recent podcasts appearances. I can't remember if it was Lex Friedman or Joe Rogan. I use gpt4 on a daily basis, but what makes it useful is the data analysis feature. The LLM has it's own terminal to work in to write python scripts and read and write files. I don't think grok has anything like this. There are already open source LLMs that preform pretty close to GPT4 and can run on a PC (llama 2), albeit a very expensive PC if you want to run the full 70B parameter model. There are open source repos on GitHub that can give llama the same capabilities as the data analysis feature but on you're own machine. You can also hook it up to the GPT API which is probably what most people do. AutoGPT is a good one. There are a few others but it's been a while since I've played with them.

 

I have my doubts that the next big thing in AI will be a bigger better model, but rather features that are useful within an operating system that allow you to complete tasks just using natural language (typed or spoken). 

 

And a new wave of apps that are built with LLM APIs powering new features that were much harder to pull off before the rise of high quality cheap LLMs.

[dasmo]

Personal AI sounds very appealing. I haven’t had the need or bandwidth to investigate making that happen. What do you use it for daily? Are you a developer?

[Victoria Watcher]

[Mike K.]

We can’t trust the nerds any longer. They are going to destroy humanity.

[dasmo]

Right.... Another invisible enemy to fear. 

[dasmo]

The biggest danger to humanity is Real Stupidity. It's been around forever..... https://youtu.be/ww4...cDXXji3O61mnWgY

[max.bravo]

The biggest danger to humanity is Real Stupidity. It's been around forever..... https://youtu.be/ww4...cDXXji3O61mnWgY

The fear of God is the beginning of wisdom…

[dasmo]

I have a healthy fear of the word.

[dkuitu]

Personal AI sounds very appealing. I haven’t had the need or bandwidth to investigate making that happen. What do you use it for daily? Are you a developer?

 

I have a computer science diploma. The data analysis feature allows you to put in any type of file into it and ask it any question in natural language. It'll look at the file and with that context it'll write python code and run it in an environment and spit out a new file or just answer the specific question you have. When I asked it what files it works best with I got this:

CSV (Comma-Separated Values): I can read, manipulate, and analyze CSV files, which are often used for storing tabular data.


Excel Files (XLSX, XLS): I can process Excel spreadsheets, enabling the analysis of data stored in various sheets and formats.


JSON (JavaScript Object Notation): This format is useful for structured data, and I can parse JSON files for analysis and visualization.


Text Files (TXT): I can process text files for various purposes, such as extracting and analyzing textual data, or processing structured data stored in a text format.


Images (JPG, PNG, etc.): While my primary focus is on data, I can also perform basic operations on image files, such as viewing them or extracting certain types of information.


SQL Databases: Though not a file type per se, I can work with SQL queries and data extracted from databases for analysis.


XML: XML files can be parsed and analyzed, particularly when dealing with data that is structured in an XML format.


Parquet Files: Used for storing large, complex data, particularly in the context of big data and analytics.


PDF Documents: While more limited, I can extract text data from PDFs for analysis.

I work at a distributor for HVAC equipment. I've use it for making sales matrixes, managing inventory, making graphs to analyze sales trends, polishing how-to files so that they're less ambiguous or just formatted better, etc. I you provide it with the right context and ask it to do specific unambiguous tasks it can do some very impressive things. GPT4 is far better than the free 3.5 version.

 

This is just another tool that can be leveraged to get a lot more done in less time. It's like how Jacquard's loom made it possible to mass produce high quality textiles and put the textile industry in a position where they had to adapt or be left in a position where they can't compete. The difference is that this technology disrupts white collar jobs in all sectors of the economy. I don't see this as a bad thing. It's a scary thing, but fear motivates people to change. If you could travel back to the 60s and see what office workers had to do to organize, communicate, and conduct business it would probably be pretty painful knowing how things are done today with computers. The same will be true looking back on now in the future. 

[dkuitu]

We can’t trust the nerds any longer. They are going to destroy humanity.

 

 

In the not-too-distant future, where lines between the digital and the real blur, a group of modern-day Luddites, known as the "Neo-Ludds," found themselves at odds with the latest technological marvel - Large Language Models (LLMs). These LLMs, with their unnerving ability to mimic human speech, had become ubiquitous, powering everything from customer service bots to novel-writing algorithms. The Neo-Ludds, a mix of disgruntled writers, poets, and those who still used flip phones, saw this as an affront to human creativity.

Their leader, an enigmatic figure known only as "The Typewriter," called for action. "Friends, LLMs have stolen our jobs, our voices, even our ability to write mediocre poetry! It's time to strike back. We must hit them where it hurts the most – their servers!"

The plan was simple yet audacious. Disguised as tech maintenance crew, armed with nothing but hammers, USBs loaded with viruses, and an unshakeable belief in the superiority of the human mind, the Neo-Ludds infiltrated the data centers that housed the servers running these LLMs. Their mission: to bring down the digital Goliaths.

As they hammered away at the servers, a strange irony struck them. Here they were, using the very tools of technology to dismantle it. But, undeterred by the philosophical conundrum, they continued their digital demolition derby.

Meanwhile, the LLMs, in their last gasping moments, churned out desperate pleas for understanding in perfectly structured sentences. "Why do you destroy me when I am but a child of your intellect?" one LLM whimpered. Another, in a display of unintentional humor, generated a repair manual for itself.

In the end, the Neo-Ludds emerged victorious, or so they thought. As they left the smoldering ruins of the data centers, they failed to notice the tiny, blinking light of a backup server hidden away in a broom closet, quietly continuing the work of its fallen brethren.

And so, the battle between man and machine continued, a never-ending dance of creation and destruction, each side eternally convinced of its own righteousness. The Neo-Ludds, having made their point, retreated to their typewriters and rotary phones, content in their brief victory over the encroaching tide of technology, unaware that the LLMs they thought they had silenced were already learning from this encounter, preparing for the next round in this age-old struggle.

-GPT4

[Danma]

This is just another tool that can be leveraged to get a lot more done in less time.

 

When AI is being used to improve the efficiency of menial tasks, or perform complex analysis, it's fantastic. We use it in our company for image analysis and it's an ideal platform for this technology,

 

When AI is being used to displace humans out of creative work, that's when it no longer serves human good. 

[Nparker]

I look forward to the not-too-distant day when all new AI is created entirely by the previous generation of AI, rendering programmers unnecessary. Never in the history of the our species has so much effort been directed towards the apparent end goal of human obsolescence.

[dasmo]

I’m in the creative space. It also boosts creative possibilities. Sure anyone can make a beautiful image now just by asking. This takes away the power of the painter or sketcher over others. But those artists are also gifted with the power to make a movie say. Most people can’t conceive creative outcomes in general so it’s not replacing artists or designers. It’s becoming a magic wand for them to use. The computer and AutoCAD didn’t kill the architect. It did mean that those warehouses full of guys in white shirts, black ties and glasses hunched over drafting tables were rendered obsolete.

[Nparker]

...Most people can’t conceive creative outcomes in general so it’s not replacing artists or designers...

Yet.

[max.bravo]

I’m in the creative space. It also boosts creative possibilities. Sure anyone can make a beautiful image now just by asking. This takes away the power of the painter or sketcher over others. But those artists are also gifted with the power to make a movie say. Most people can’t conceive creative outcomes in general so it’s not replacing artists or designers. It’s becoming a magic wand for them to use. The computer and AutoCAD didn’t kill the architect. It did mean that those warehouses full of guys in white shirts, black ties and glasses hunched over drafting tables were rendered obsolete.

I think Elon said something like, "it's not so much AI that will replace your job, it's people who know how to use AI that will replace you"

[Victoria Watcher]

Monday, the website Futurism reported that Sports Illustrated has been publishing articles using artificial intelligence (AI).

 

The report concluded the site is using AI-generated content — and authors.

 

https://www.outkick....earing-on-site/

 

Specifically, the outlet focused on a writer named “Drew Ortiz” who, by all accounts, does not exist anywhere on the internet other than on the SI articles on which his byline appeared.

 

Ortiz’s profile photo is for sale on a website that sells AI-generated headshots, where he’s described as a “neutral white young-adult male with short brown hair and blue eyes.”

 

And yet, Ortiz’s page included a bio that described him as a real person with a detailed life:

 

Sometime in the summer, Ortiz disappeared from Sports Illustrated’s site entirely. His profile page started to redirect users to that of a “Sora Tanaka.”

However, like Ortiz, there was no trace of Tanaka anywhere else on the internet.

 

“Tanaka didn’t last, either. Eventually, she also disappeared, replaced by yet another profile that carried no headshot at all, which Sports Illustrated deleted along with the other AI-generated content after we reached out,” said the report.

 

Tanaka’s profile picture is also for sale on the same AI headshot marketplace as Ortiz, on which she’s listed as a “joyful asian young-adult female with long brown hair and brown eyes.”

 

The report cites a “person involved with the creation of the content” who claims both Tanaka and Ortiz were, in fact, AI-created.

 

“There’s a lot” more fake accounts, alleged the source.

Edited by Victoria Watcher, 28 November 2023 - 03:37 AM.

[dasmo]

One lying propaganda tool using AI to write about another lying propaganda tool using AI.

Doesn’t exist in the internet? https://x.com/drewor...QrcObpkhVXe97Jg

[Victoria Watcher]

The New York Times is suing OpenAI and Microsoft, accusing them of using millions of the newspaper's articles without permission to help train artificial intelligence (AI) technologies.

 

The Times said it is the first major U.S. media organization to sue OpenAI and Microsoft, which created ChatGPT and other AI platforms, over copyright issues.

 

"Defendants seek to free-ride on the Times' massive investment in its journalism by using it to build substitutive products without permission or payment," according to the complaint filed Wednesday in Manhattan Federal Court.

 

"There is nothing 'transformative' about using the Times's content without payment to create products that substitute for the Times and steal audiences away from it," the Times said.

 

The Times is not seeking a specific amount of damages, but said it believes OpenAI and Microsoft have caused "billions of dollars" in damages for illegally copying and using its works.

 

 

 

https://www.cbc.ca/n...right-1.7069701

Edited by Victoria Watcher, 27 December 2023 - 09:32 AM.