HIGH-INCOME SKILLS TO LEARN๐ฐ
1. Artificial Intelligence
2. Cloud Computing
3. Data Science
4. Machine Learning
5. Blockchain
6. Data Analytics
7. Data Engineering
8. Applications Engineering
9. Web Development
10. Software Development
11. UX Design
12. Web Design
13. Graphic Design
14. Video Editing
15. Content Marketing
16. Digital Marketing
1. Artificial Intelligence
2. Cloud Computing
3. Data Science
4. Machine Learning
5. Blockchain
6. Data Analytics
7. Data Engineering
8. Applications Engineering
9. Web Development
10. Software Development
11. UX Design
12. Web Design
13. Graphic Design
14. Video Editing
15. Content Marketing
16. Digital Marketing
๐7โค4
OpenAI New Model-01
Don't let snake-oil salesmen fool you. This new model released by OpenAI today doesn't "think." It just generates an extensive "chain of thought", which is just a discussion of the model with itself that looks like a person talking to themselves.
Previous models were trained to be used in one-shot. You ask a question, you get an answer. Because of randomness in next token generation, if you were unlucky, your answer might be wrong. This model was finetuned to generate a long (and hidden from the user by the UI) discussion on how to better solve the problem, what facts are known, what assumptions need to be made, and what constraints should be respected.
If you explicitly asked previous models to generate this discussion before answering your question, you would get a better quality result, because the final answer would be conditioned on the information contained in this discussion.
They seemingly optimized their model to generate good quality discussions (without the user asking for it) by using reinforcement learning on various problems that have a verifiable solution, so that a reward for finding the right answer could be automatically assigned. For example:
Question: 1+1 = ?
Discussion: we have 1 and we have 1 more. And we have a plus sign, so it's an addition. What happens if we add 1 and 1? It means 1 is incremented by 1. When we increment 1 by 1, what do we get? Let's count: 1, 2, 3, 4,... Ok, 2 comes after 1, so 1 + 1 must be 2.
Answer: 2
Reward: 1
Question: 1+1 = ?
Discussion: It's easy. 1+1=11
Answer: 11
Reward: 0
Once the model is trained, what the user sees:
Question: 1+1 = ?
(Discussion happens behinds the scenes.)
Answer: 2
Sure! Here's a more polished version of the statement:
Don't get swept away by the hype around AI; Stay grounded and approach it thoughtfully. ๐ฏ
Don't let snake-oil salesmen fool you. This new model released by OpenAI today doesn't "think." It just generates an extensive "chain of thought", which is just a discussion of the model with itself that looks like a person talking to themselves.
Previous models were trained to be used in one-shot. You ask a question, you get an answer. Because of randomness in next token generation, if you were unlucky, your answer might be wrong. This model was finetuned to generate a long (and hidden from the user by the UI) discussion on how to better solve the problem, what facts are known, what assumptions need to be made, and what constraints should be respected.
If you explicitly asked previous models to generate this discussion before answering your question, you would get a better quality result, because the final answer would be conditioned on the information contained in this discussion.
They seemingly optimized their model to generate good quality discussions (without the user asking for it) by using reinforcement learning on various problems that have a verifiable solution, so that a reward for finding the right answer could be automatically assigned. For example:
Question: 1+1 = ?
Discussion: we have 1 and we have 1 more. And we have a plus sign, so it's an addition. What happens if we add 1 and 1? It means 1 is incremented by 1. When we increment 1 by 1, what do we get? Let's count: 1, 2, 3, 4,... Ok, 2 comes after 1, so 1 + 1 must be 2.
Answer: 2
Reward: 1
Question: 1+1 = ?
Discussion: It's easy. 1+1=11
Answer: 11
Reward: 0
Once the model is trained, what the user sees:
Question: 1+1 = ?
(Discussion happens behinds the scenes.)
Answer: 2
Sure! Here's a more polished version of the statement:
Don't get swept away by the hype around AI; Stay grounded and approach it thoughtfully. ๐ฏ
๐12
In my teen's I thought:
โข A fancy job title.
โข A huge salary.
โข A demanding job.
= WINNING
In my 20s I realized:
โข Having control of your time.
โข Doing work you enjoy.
โข Seeing family more happy.
= WINNING
Life Looks Different Every Decade.
โข A fancy job title.
โข A huge salary.
โข A demanding job.
= WINNING
In my 20s I realized:
โข Having control of your time.
โข Doing work you enjoy.
โข Seeing family more happy.
= WINNING
Life Looks Different Every Decade.
๐20โค1๐1
โโโโ๐ How to generate a photo of a non-existent person! ๐
๐ If you want to create a fake account on a social network, you can use another person's photo, but this is not the best option. It is better to use the following service to generate photos of non-existent people:
๐คฏ. Open this website: https://thispersondoesnotexist.com/
๐คฏ. Visiting the website, we immediately get a photo of a non-existent person.
๐คฏ. Updating the page, you will see a new generated image.
โ ๏ธ That's it, you can update the resource until you are satisfied with the photo. The site works very fast which is an undoubted plus. Many sites based on the work of artificial intelligence are often very slow. โ ๏ธ
โก๏ธ Need 200 Reactions on this Post
๐ If you want to create a fake account on a social network, you can use another person's photo, but this is not the best option. It is better to use the following service to generate photos of non-existent people:
๐คฏ. Open this website: https://thispersondoesnotexist.com/
๐คฏ. Visiting the website, we immediately get a photo of a non-existent person.
๐คฏ. Updating the page, you will see a new generated image.
โ ๏ธ That's it, you can update the resource until you are satisfied with the photo. The site works very fast which is an undoubted plus. Many sites based on the work of artificial intelligence are often very slow. โ ๏ธ
โก๏ธ Need 200 Reactions on this Post
๐15
FREE BOOKS
+==========+
Download programming books for FREE, all books from 2019!
https://t.iss.one/progerbooks
+==========+
Download programming books for FREE, all books from 2019!
https://t.iss.one/progerbooks
๐4
We are now on WhatsApp as well
Follow for more Artificial Intelligence resources: ๐
https://whatsapp.com/channel/0029VaoePz73bbV94yTh6V2E
Follow for more Artificial Intelligence resources: ๐
https://whatsapp.com/channel/0029VaoePz73bbV94yTh6V2E
Unpopular opinion:
ChatGPT is only as smart as the user; if garbage goes in, garbage comes out.
ChatGPT is only as smart as the user; if garbage goes in, garbage comes out.
๐9โค4
๐ ๐๐ฎ๐ฌ๐ญ-๐๐๐ญ๐๐ก ๐๐ ๐๐๐ ๐๐๐ฅ๐ค๐ฌ
โฉ The inside story of ChatGPT's astonishing potential by Greg Brockman. https://youtu.be/C_78DM8fG6E?si=kdGNA1PvO1lb7L8t
โฉ How AI could save (not destroy) education by Sal Khan
https://youtu.be/hJP5GqnTrNo?si=wlD-SOjr5ZxLQ0vQ
โฉ How to keep AI under control by Max Tegmark
https://youtu.be/xUNx_PxNHrY?si=e8JDz9up3IRYmBo5
โฉ How to think computationally about AI, the universe, and everything by Stephen Wolfram
https://youtu.be/fLMZAHyrpyo?si=5O1b63qgga89rEOb
โฉ The dark side of competition in AI by Liv Boeree
https://youtu.be/WX_vN1QYgmE?si=QDMlKkrxqrSCdFkr
โฉ How AI art could enhance humanity's collective memory by Refik Anadol
https://youtu.be/iz7diOuaTos?si=iyQOF20jZp78hfo2
โฉ Why AI is incredibly smart and shockingly stupid by Yejin Choil
https://youtu.be/SvBR0OGT5VI?si=rLhDzmohC_dPfrtM
โฉ Will superintelligent AI end the world by Eliezer Yudkowsky
https://youtu.be/Yd0yQ9yxSYY?si=JqN2yNgP0IOTnjN1
โฉ Kai-Fu Lee: How AI can save our humanity | TED Talk
https://bit.ly/44SC1ss
โฉ The inside story of ChatGPT's astonishing potential by Greg Brockman. https://youtu.be/C_78DM8fG6E?si=kdGNA1PvO1lb7L8t
โฉ How AI could save (not destroy) education by Sal Khan
https://youtu.be/hJP5GqnTrNo?si=wlD-SOjr5ZxLQ0vQ
โฉ How to keep AI under control by Max Tegmark
https://youtu.be/xUNx_PxNHrY?si=e8JDz9up3IRYmBo5
โฉ How to think computationally about AI, the universe, and everything by Stephen Wolfram
https://youtu.be/fLMZAHyrpyo?si=5O1b63qgga89rEOb
โฉ The dark side of competition in AI by Liv Boeree
https://youtu.be/WX_vN1QYgmE?si=QDMlKkrxqrSCdFkr
โฉ How AI art could enhance humanity's collective memory by Refik Anadol
https://youtu.be/iz7diOuaTos?si=iyQOF20jZp78hfo2
โฉ Why AI is incredibly smart and shockingly stupid by Yejin Choil
https://youtu.be/SvBR0OGT5VI?si=rLhDzmohC_dPfrtM
โฉ Will superintelligent AI end the world by Eliezer Yudkowsky
https://youtu.be/Yd0yQ9yxSYY?si=JqN2yNgP0IOTnjN1
โฉ Kai-Fu Lee: How AI can save our humanity | TED Talk
https://bit.ly/44SC1ss
๐5โค2
A resent post from CEO of Telegram about the new policies ๐
๐ Search on Telegram is more powerful than in other messaging apps because it allows users to find public channels and bots. Unfortunately, this feature has been abused by people who violated our Terms of Service to sell illegal goods.
๐ช Over the last few weeks, a dedicated team of moderators, leveraging AI, has made Telegram Search much safer. All the problematic content we identified in Search is no longer accessible. If you still manage to find something unsafe or illegal in Telegram Search, please report it to us via @SearchReport.
๐ซ To further deter criminals from abusing Telegram Search, we have updated our Terms of Service and Privacy Policy, ensuring they are consistent across the world. Weโve made it clear that the IP addresses and phone numbers of those who violate our rules can be disclosed to relevant authorities in response to valid legal requests.
โ๏ธ These measures should discourage criminals. Telegram Search is meant for finding friends and discovering news, not for promoting illegal goods. We won't let bad actors jeopardize the integrity of our platform for almost a billion users.
๐ Search on Telegram is more powerful than in other messaging apps because it allows users to find public channels and bots. Unfortunately, this feature has been abused by people who violated our Terms of Service to sell illegal goods.
๐ช Over the last few weeks, a dedicated team of moderators, leveraging AI, has made Telegram Search much safer. All the problematic content we identified in Search is no longer accessible. If you still manage to find something unsafe or illegal in Telegram Search, please report it to us via @SearchReport.
๐ซ To further deter criminals from abusing Telegram Search, we have updated our Terms of Service and Privacy Policy, ensuring they are consistent across the world. Weโve made it clear that the IP addresses and phone numbers of those who violate our rules can be disclosed to relevant authorities in response to valid legal requests.
โ๏ธ These measures should discourage criminals. Telegram Search is meant for finding friends and discovering news, not for promoting illegal goods. We won't let bad actors jeopardize the integrity of our platform for almost a billion users.
๐4
๐ง๐ก Why is OpenAI Paying Publishers if it Already Took Their Data? ๐ป
OpenAI has been quietly striking deals with major publishers, even after scraping their content without permission to train its models. But why pay now? The answer lies in legal strategy and future-proofing its AI tools like ChatGPT. By securing these agreements, OpenAI avoids lawsuits like the one from The New York Times, which could cost them billions in damages.
These deals also give OpenAI access to more real-time information, enhancing ChatGPTโs relevance and accuracy. ๐ฐ The catch? Publishers get peanutsโ$1-10 million per yearโcompared to the huge value OpenAI extracts from their data.
In essence, these are preemptive settlements to avoid further legal headaches while maintaining a competitive edge against rivals like Google. ๐ With its new SearchGPT, OpenAI is eyeing the search engine space, hoping to outmaneuver Googleโs dominance. ๐ต๏ธโโ๏ธ
This is more than just paying for past sins; itโs about shaping the future of AI search.
OpenAI has been quietly striking deals with major publishers, even after scraping their content without permission to train its models. But why pay now? The answer lies in legal strategy and future-proofing its AI tools like ChatGPT. By securing these agreements, OpenAI avoids lawsuits like the one from The New York Times, which could cost them billions in damages.
These deals also give OpenAI access to more real-time information, enhancing ChatGPTโs relevance and accuracy. ๐ฐ The catch? Publishers get peanutsโ$1-10 million per yearโcompared to the huge value OpenAI extracts from their data.
In essence, these are preemptive settlements to avoid further legal headaches while maintaining a competitive edge against rivals like Google. ๐ With its new SearchGPT, OpenAI is eyeing the search engine space, hoping to outmaneuver Googleโs dominance. ๐ต๏ธโโ๏ธ
This is more than just paying for past sins; itโs about shaping the future of AI search.
๐Top Sites To Explore Public Files - 2024
Itโs the internet equivalent of exploring an old attic - but instead of a flashlight, youโre equipped with keywords.
๐ขFind public files in Amazon S3
๐ฅ Website: ๐https://s3digger.com
๐ขFind public files in Dropbox
๐ฅ Website: ๐https://drodigger.com
๐ขFind public files in Disk Yandex
๐ฅ Website: ๐https://yadigger.com
๐ขFind public files in Google Drive
๐ฅ Website: ๐https://drodigger.com
๐ขFind public files in Telegram Server
๐ฅ Website: ๐https://teledigger.com
๐ขFind public files in other file sharing sites
๐ฅ Website: ๐https://fidigger.com
๐ขFind public files on url shortening sites
๐ฅ Website: ๐https://shortdigger.com
Itโs the internet equivalent of exploring an old attic - but instead of a flashlight, youโre equipped with keywords.
๐ขFind public files in Amazon S3
๐ฅ Website: ๐https://s3digger.com
๐ขFind public files in Dropbox
๐ฅ Website: ๐https://drodigger.com
๐ขFind public files in Disk Yandex
๐ฅ Website: ๐https://yadigger.com
๐ขFind public files in Google Drive
๐ฅ Website: ๐https://drodigger.com
๐ขFind public files in Telegram Server
๐ฅ Website: ๐https://teledigger.com
๐ขFind public files in other file sharing sites
๐ฅ Website: ๐https://fidigger.com
๐ขFind public files on url shortening sites
๐ฅ Website: ๐https://shortdigger.com
๐8
How to search on Google from another region
I Search From helps you simulate using a Google search from another location or device, or search with custom search settings.
The site allows you to build your search query so that Google searches by the territorial data you specify.
You can choose from search region and device type to city.
#lifehack #security
I Search From helps you simulate using a Google search from another location or device, or search with custom search settings.
The site allows you to build your search query so that Google searches by the territorial data you specify.
You can choose from search region and device type to city.
#lifehack #security
โค1
16 Websites to Find Remote International Jobs
1. LinkedIn - linkedin.com
2. Indeed - indeed.com
3. Glassdoor - glassdoor.com
4. FlexJobs - flexjobs.com
5. Remote.co - remote.co
6. Upwork - upwork.com
7. Freelancer - freelancer.com
8. Fiverr - fiverr.com
9. Guru - guru.com
10. Toptal - toptal.com
11. AngelList - angel.co
12. SimplyHired - simplyhired.com
13. Remotive - remotive.com
14. Hired - hired.com
15. CloudPeeps - cloudpeeps.com
16. TaskRabbit - taskrabbit.com
1. LinkedIn - linkedin.com
2. Indeed - indeed.com
3. Glassdoor - glassdoor.com
4. FlexJobs - flexjobs.com
5. Remote.co - remote.co
6. Upwork - upwork.com
7. Freelancer - freelancer.com
8. Fiverr - fiverr.com
9. Guru - guru.com
10. Toptal - toptal.com
11. AngelList - angel.co
12. SimplyHired - simplyhired.com
13. Remotive - remotive.com
14. Hired - hired.com
15. CloudPeeps - cloudpeeps.com
16. TaskRabbit - taskrabbit.com
๐13๐ฅฐ1
10 awesome frontend development YouTube channels:
1. Traversy Media ๐
2. The Net Ninja ๐ฅท
3. Dev Ed ๐จ
4. Academind ๐
5. Fireship ๐ฅ
6. Codevolution ๐ป
7. DesignCourse ๐จ
8. Florin Pop ๐งโ๐ป
9. Web Dev Simplified ๐
10. Kevin Powell ๐ฅ
@Javascript_courses
1. Traversy Media ๐
2. The Net Ninja ๐ฅท
3. Dev Ed ๐จ
4. Academind ๐
5. Fireship ๐ฅ
6. Codevolution ๐ป
7. DesignCourse ๐จ
8. Florin Pop ๐งโ๐ป
9. Web Dev Simplified ๐
10. Kevin Powell ๐ฅ
@Javascript_courses
โค6๐2
๐ New Telegram Channel Alert!
We've discovered a gem for AI enthusiasts: AI.News.Daily!
Why follow AI.News.Daily?
It's your go-to for:
โข Daily AI Breakthroughs: Get updates every day.
โข Expert Insights: Learn from the best with easy-to-follow advice.
โข Monetization Tips: Discover how to turn AI into profit.
Top picks from the channel:
โข Anthropicโs Free AI Courses: Access a treasure trove of free courses and guides tailored for Claude.
โข AI Catalogs: Explore the top 10 AI catalogs on GitHub.
โข Free Prompt Libraries: Stay creative with the 6 best AI prompt libraries.
โข Website Building: Discover top AI solutions for creating websites effortlessly.
โข AI Selection Tool: Compare AIs, check prices, and view detailed graphs and leaderboards.
Thereโs so much more waiting for you!
๐Donโt miss out: @ainews_daily
We've discovered a gem for AI enthusiasts: AI.News.Daily!
Why follow AI.News.Daily?
It's your go-to for:
โข Daily AI Breakthroughs: Get updates every day.
โข Expert Insights: Learn from the best with easy-to-follow advice.
โข Monetization Tips: Discover how to turn AI into profit.
Top picks from the channel:
โข Anthropicโs Free AI Courses: Access a treasure trove of free courses and guides tailored for Claude.
โข AI Catalogs: Explore the top 10 AI catalogs on GitHub.
โข Free Prompt Libraries: Stay creative with the 6 best AI prompt libraries.
โข Website Building: Discover top AI solutions for creating websites effortlessly.
โข AI Selection Tool: Compare AIs, check prices, and view detailed graphs and leaderboards.
Thereโs so much more waiting for you!
๐Donโt miss out: @ainews_daily
โค9๐5