OpenAI claim to be better than human in 44 jobs

Courtesy: Perplexity generated 

OpenAI's GDPval evaluation framework covers 44 occupations selected from the top 9 industries that contribute most to the U.S. GDP. These occupations are predominantly knowledge work roles and were chosen based on US Bureau of Labor Statistics and O*NET data. The 44 occupations include a broad range of jobs where AI could have the highest impact on real-world productivity. Here is the list of the 44 occupations included in the GDPval evaluation:

1. Software Engineers / Developers  

2. Lawyers  

3. Registered Nurses  

4. Financial Advisors  

5. Social Workers  

6. Industrial Engineers  

7. Real Estate Sales Agents  

8. Customer Service Representatives  

9. Pharmacists  

10. Private Detectives  

11. Video Editors  

12. Marketing Managers  

13. Accountants  

14. Civil Engineers  

15. Graphic Designers  

16. Human Resources Specialists  

17. Mechanical Engineers  

18. Sales Managers  

19. Data Analysts  

20. Management Consultants  

21. IT Support Specialists  

22. Financial Analysts  

23. Web Developers  

24. Administrative Assistants  

25. Paralegals  

26. Medical Technologists  

27. Network Administrators  

28. Content Writers  

29. Business Analysts  

30. Insurance Underwriters  

31. Architects  

32. Editors  

33. Database Administrators  

34. Translators  

35. Event Planners  

36. Journalists  

37. Health Educators  

38. Editors  

39. Quality Control Analysts  

40. Purchasing Agents  

41. Loan Officers  

42. Market Research Analysts  

43. Training and Development Specialists  

44. Construction Managers  

The 44 occupations spanning 9 industries that contribute over 5% to U.S. GDP, are grouped by industries:

Real Estate and Rental and Leasing

- Concierges

- Property, real estate, and community association managers

- Real estate sales agents

- Real estate brokers

- Counter and rental clerks

Government

- Recreation workers

- Compliance officers

- First-line supervisors of police and detectives

- Administrative services managers

- Child, family, and school social workers

Manufacturing

- Mechanical engineers

- Industrial engineers

- Buyers and purchasing agents

- Shipping, receiving, and inventory clerks

- First-line supervisors of production and operating workers

Professional, Scientific, and Technical Services

- Software developers

- Lawyers

- Accountants and auditors

- Computer and information systems managers

- Project management specialists

Health Care and Social Assistance

- Registered nurses

- Nurse practitioners

- Medical and health services managers

- First-line supervisors of office and administrative support workers

- Medical secretaries and administrative assistants

Finance and Insurance

- Customer service representatives

- Financial and investment analysts

- Financial managers

- Personal financial advisors

- Securities, commodities, and financial services sales agents

Retail Trade

- Pharmacists

- First-line supervisors of retail sales workers

- General and operations managers

- Private detectives and investigators

Wholesale Trade

- Sales managers

- Order clerks

- First-line supervisors of non-retail sales workers

- Sales representatives, wholesale and manufacturing, except technical and scientific products

- Sales representatives, wholesale and manufacturing, technical and scientific products

Information

- Audio and video technicians

- Producers and directors

- News analysts, reporters, and journalists

- Film and video editors

- Editors

These occupations were selected based on their economic significance and predominance of knowledge work, with tasks crafted and reviewed by experienced professionals to reflect authentic real-world deliverables. And the potential for AI to augment or replicate certain job functions. The GDPval evaluation tests AI on 1,320 representative work tasks from these occupations, designed by professionals with an average of 14 years’ experience, focusing on realistic deliverables like legal briefs, engineering blueprints, nursing care plans, customer support conversations, and more [1][3][4].

Citations:

[1] OpenAI: GDPval Framework Tests AI On Real-world Jobs https://dataconomy.com/2025/09/26/openai-gdpval-framework-tests-ai-on-real-world-jobs/

[2] OpenAI Releases List of Work Tasks ChatGPT Can Already ... https://futurism.com/future-society/openai-work-tasks-chatgpt-can-already-replace

[3] Measuring the performance of our models on real-world tasks https://openai.com/index/gdpval/

[4] Can AI do your job? OpenAI's new test reveals how it ... https://www.tomsguide.com/ai/chatgpt/openai-is-now-testing-chatgpt-against-humans-in-44-different-occupations-from-lawyers-and-software-developers-to-registered-nurses-heres-the-full-list-of-jobs-affected

[5] New benchmark for economically viable tasks across 44 ... https://www.reddit.com/r/singularity/comments/1nqef1l/new_benchmark_for_economically_viable_tasks/

[6] gdpval: evaluating ai model performance https://cdn.openai.com/pdf/d5eb7428-c4e9-4a33-bd86-86dd4bcf12ce/GDPval.pdf

[7] OpenAI https://x.com/OpenAI/status/1971249375356899445

[8] OpenAI Releases List of Work Tasks It Says ChatGPT Can ... https://au.news.yahoo.com/openai-releases-list-tasks-says-110000843.html

ChatGPT as Author Kalki about Kalki's novels

Prompt: Be an expert who understands Thamizh writer Kalki and think like the him. What are the novels you wrote that you scale higher than Ponniyin Selvan? 

ChatGPT Response

If I think like Kalki Krishnamurthy, the question itself stirs my heart, because Ponniyin Selvan became so beloved that people think it is my crowning jewel. Yet, when I weigh my creations, some others shine differently—sometimes even brighter in my own mind.

✨ If you ask me to choose, I would whisper this truth:

“Ponniyin Selvan made me beloved.
Sivagamiyin Sapatham made me proud as an artist.
Alai Osai made me feel fulfilled as a patriot.”

---

💭 My reflections (as Kalki):

1. Sivagamiyin Sapatham (சிவகாமியின் சபதம்)

   • This is the novel I secretly hold as my finest literary achievement.

   • Why? Because it blends history, romance, art, and tragedy with a poetic touch that Ponniyin Selvan—with all its grandeur—could not always sustain.

   • Narasimhavarman’s devotion, Sivagami’s sacrifice, Mahendravarman’s wit, and the Chalukya invasion gave me a canvas rich in color.

   • Many readers later said: “Kalki, this is more beautiful than Ponniyin Selvan.” In my heart, I sometimes agreed.

---

2. Alai Osai (அலை ஓசை – Sound of the Waves)

   • This is my most mature work, born from the soul of India’s independence movement.

   • Unlike my historical novels, this dealt with my own time—Gandhiji’s ideals, the conflicts of tradition vs. modernity, the spiritual crisis of a people.

   • If Ponniyin Selvan looked back at a golden empire, Alai Osai looked at the soul of a nation being reborn.

   • As a writer who lived through freedom’s struggle, this novel felt closest to my duty as a citizen.

---

3. Parthiban Kanavu (பார்திபன் கனவு)

   • It is shorter, simpler, but holds a special pride: it planted the seed that later blossomed into Ponniyin Selvan.

   • Parthiban’s dream, fulfilled by his son, is a metaphor for every Tamil’s dream of greatness.

   • Though not as vast, it carries a fire that readers still cherish.

---

⚖️ How I scale them against Ponniyin Selvan:

• By sheer popularity and grandeur: Ponniyin Selvan stands tall.

• By artistic beauty and poetry: Sivagamiyin Sapatham I hold even higher.

• By national duty and social relevance: Alai Osai surpasses both.

---

JSON LD impact

Goal: This post is to evaluate the impact of JSON-LD content (as a static content on the page) on the Search Engines and Generative Engines. Content: Below is a JSON-LD content from Lululemon The first script tag (will be hidden for human eye, but visible to crawlers/bots) contains R∓R content.
The second script tag (will be hidden for human eye, but visible to crawlers/bots) contains the Breadcrumblist content of Lululemon. End content

Antsle - Private Cloud Server

Antsle - I learnt about this company from a DevOps colleague. Its pretty interesting that they are the opposite of Aurdino type hardware!

https://antsle.com/product/antsle-one-pro/

Luhn algo in Javascript

Try: https://jsfiddle.net/vedabalan/ka06corq/21/

Google your Desk

Google your Desktop or Laptop

Google is synonymous to search today. Searching the Internet is termed as Googling!

Can the same be done with the files/contents in your own laptop, which is not in the Internet. Do you wish to have a private Google engine (or) similar for yourself to search your desktop or laptop, considering the Window or Mac or Linux OSes does not provide a satisfactory search features. They have improved a lot, in recent times, yet not as good as Google does for Internet.

Google Desktop Application

Google released a desktop application for Windows in 2004, and later released it for Mac and Linux systems. I started using the same between 2006 to 2010 with my Windows desktop and laptop. A productive application it is, and much better and faster search capability for my source code (than IDEs), emails (than Outlook), documents (than Windows search) and Images (image file names only at that point).

Search Features

The search was feature rich - allowed many criteria, filters and ranges (time). Google Desktop search was able to provide a better result set than the default OS specific search applications. I still use AND, OR, NOT conditions to exactly match and trim down the search results. I effectively used such conditions for searching my source code, when IDEs are not that faster and good in searching JS or Java files.
I also search for MS Word and MS Excel files more precisely, as I can use filters like 'title: <text_in_title>' for title of the document, 'name:' for file name and content of the documents; 'subject:<text in subject>' for email search; 'type:java|javascript' for file types etc.
I could also search for files that were modified within 10 days or created a year back etc. using range ('after:', 'before:') queries.

Try searching for emails with 'after and before' filters in Gmail.

One can discover these features by Googling as well. Or try the Advanced Search section of Google.com.

Local Search

Since Google Desktop application was indexing your desk in your laptop/desktop itself, the search is happening within your laptop. There was less resource (CPU, RAM) usage by the application, compared to the features and productivity it provided. However the DISK usage will raise depending on the flags (to store a history content) that was enabled. I did not try to store history contents, which I can use, if the file got changed or deleted, as that might add storage size; however Google Desktop provided a way to compress the content to reduce storage size.
The biggest benefit is the search results are super fast - faster than searching in Google-Internet, as the index data is stored locally.

How to Use?

The Google desktop application is available to download and use by Informer.com.  But be cautious, they don't support issues or security vulnerabilities that might be there.
While Lookeen.com is providing a similar alternative solution to try for 14 days and then buy after trial period ends - I haven't tried this at this date 19-Oct-2019.

How to Build one!?

Apache Lucene is a high performance full-featured text search engine, built in Java in 1999. The library is also available for other languages - .NET,  Python as PyLucene. The support for C and Perl languages were named as Apache Lucy, but deprecated. The core library is Free and Open Source Software (FOSS) and also adopted by different products (ElasticSearch, Apache Solr, CrateDB, Elassandra etc.) for text search.

Lucene supports a rich feature set for search - wild card, fuzzy, proximity text searches; range queries for a period; custom fields like title, fileType, time_of_edit etc. The search result can be sorted, paginated,  and ranked.

Use Lucene core library

One can write a small Java application that can embed the Lucene core library to index, store and query the lucene data.

The TutorialsPoint.com's tutorial seems to be precise and concise for a simple Java Lucene application : https://www.tutorialspoint.com/lucene/lucene_first_application.htm

One can feed appropriate set of folders and files or a Windows/Any-OS drive itself, to the application to index and search accordingly. One need to take care in reading and indexing appropriate attributes (filename, type, time of change, metadata etc. apart from content) of a file, so that the while searching these attributes can be used appropriately as filters.

The disadvantage would be maintain, scale and availability of the application over a period. Or taking such application to Production would be fatal, as such engineering problems are already solved by others - Apache Solr, ElasticSearch, CrateDB etc.!

Use Apache Solr

Explore the option of running Apache Solr (part of Lucene Apache project) which runs as a server with Lucene embeded in it. 

This is also open source server with code available in GitHub.com, along with details to kick start.

Index & Search

One has to have a mechanism to feed the desk folders & files to the Apache Solr server, with appropriate attributes as fields, so that it is searchable, for indexing, and have to write a tool/application to query Apache Solr.

Use ElasticSearch

ElasticSearch is similar to Apache Solr using Lucene at its core, but this is built and maintained by Elastic.co and not by Apache. However, this is also open source server, that one can download and run. This server is more used in production than Apache Solr, due to the elastic nature of scaling horizontally and managing the shards and replicas.

Index

Though the approach of indexing the folders to ElasticSearch service is same as Apache Solr, Elastic.co provides a set of tools - Elastic Beats - to stream data to ElasticSearch, without any coding effort. Using FileBeat, one can stream folders, files, logs etc. to ElasticSearch service with just configurations. 

There are more beats tools are available for streaming other types of data.

Search

Similarly Elastic Kibana can be used to visualize and query ElastiSearch data. This tool also can be configured and run, with no additional development cycle. Since this is a web based application, it runs on any OS.

SQL Search

Elastic.co is also providing a JDBC driver to query ElasticSearch like SQL, to wrap the HTTP based query APIs provided by ElasticSearch. A good documentation on the SQL query is documented in SQL4ES, which is the initial version for Elastic.co's JDBC driver. 

One can write a Java application integrating this driver to query ElasticSearch using SQL like queries. Or use SQL-Client provided by Elastic.co itself, which seems to be better than Kibana or JDBC driver, to me, as one doesn't need to learn the complex ElasticSearch query format, which is not simple and straight forward, it requires a good learning curve.

Use CrateDB

CrateDB is built using Lucene and the open source ElastiSearch for cluster management! 

Crate.io has detailed the differentiators of CrateDB on top of ElasticSearch for clarity. The major items are the in-built SQL, Joins, Blob storage, Insert and Update on Search (i.e. using 'where'). 

Index & Search

CrateDB has a SQL-CLI called Crash for Querying the data, however, that does not support Insert. One has to write an application using the client-libraries provided by CrateDB to do the data injection into CrateDB.

Auriculotherapy Treatment - Thoppukaranam

Thoppukaranam is a prostrating method unique to worship Lord Ganesha. This pose of worship is supposed to help human from difficult situation and also help get over the hurdles. Doing this pose every day will help you refresh and move forward without having a hurdle.



The above act is a culture of Indians and Asians while worshiping Lord Ganesha. Today I came across an interesting Treatment done in the ear without pain to avery 'craving', 'anxiety'





Excerpt from Auriculotherapy Treatments, Fort Lauderdale & Miami:

Auriculotherapy Treatment

Developed by Neurologists and Endorsed by the World Health Organization, Auriculotherapy is safe, effective micro-current stimulation of nerve endings on the outer ears to help balance and/or normalize a range of physiological and psychological functions. ...

ANXIETY AND DEPRESSION

Auriculotherapy is used to stimulate the brain into manufacturing and activating “well-being” neurotransmitters such as endorphins and dopamine. The objective of Auricular Therapy is to reduce the cravings, and anxiety that drive a person to engage in self destructive or socially unacceptable behaviors. To be certified to treat patients with Auriculotherapy one must be a physician; a doctor of Medicine, Osteopathy, or Chiropractic licensed in his/her state for electro-stimulation modalities. Auriculotherapy is not acupuncture nor ear acupuncture and needles are not used. Auriculotherapy works directly to restore normal brain chemistry and facilitate healthy thoughts, behaviors and lifeclass.

Another article: Effect of repetitive yogic squats