AI teamwork: Using the output of ChatGPT to generate DALL-E 2 art.

What would happen if I chained two AIs together to work as a team? ChatGPT is a conversationAI at chat.openai.com that can generate answers to questions or produce detailed descriptions given minimal input. DALL-E 2 creates art from elaborate textual descriptions.

So I asked ChatGPT to:

Describe a business scene with people collaborating on a piece of art.

It responded with:

In a bustling art studio, a team of talented artists are collaborating on a large-scale piece of art. They are all working together, sharing ideas and techniques, and using their individual strengths to bring the piece to life. Some are sketching out rough drafts and ideas on paper, while others are working on the more technical aspects of the piece, such as adding color or texture. There is a sense of excitement and creativity in the air, as the team works together to create something truly unique and beautiful. As the piece starts to take shape, the artists share their thoughts and feedback with each other, making tweaks and adjustments along the way to ensure that the final product is exactly what they envision. In the end, the team is able to create a stunning piece of art that they are all proud of, thanks to their collaboration and dedication to the project.

I pushed this output, at least what would fit, into DALL-E 2 and it generated a few options. You can be the judge of how well these images meet the original request. With DALL-E 2, you can ask for variations of each of the original images to continue exploring for the image that fits the bill. As to “business scene” in the original request, I think the third image below come closest, although I do find the appearance of the craftsman’s apron fascinating in the second image. The implications for the future are astounding for AI as a creative tool.

Sometimes you gotta go old school

Yes, the handset works!

I forgot my high-tech Bluetooth headphones and had to reach for this handset. The sound quality is actually amazing and there is a whole additional benefit. People (at least Gen-X and boomers…and my kids!) get that I’m actually on the phone and talking to someone not in the room. Sometimes, this old tech provides the ambient information to others that is lost in our continually connected, wireless world. Combine with this with the Lightning to 3.5 mm Audio dongle from Apple, and you can use this with your iPhone, too. It works really well, can leave your phone in the pocket and people will know you are on the phone, even while walking down the street.

Looking for your headphones, but they are too tiny to find? This big handset is much harder to lose under the bed, in the covers, or under the car seat.

If you are looking for that On-Air indicator or looking for a way to notify others and let them know you are on a call or on the phone, I highly recommend picking up one of these today!

Traveling the world and trying to npm a webapp?

How To Fix NPM Timeout Errors On Slow Internet Connections

Is npx create-react-app failing with timeouts while you are on a good old fashioned copper connection?

Sometimes when we work with slow or unstable Internet connections, npm often fails to complete its commands, like npm install and npm update with a timeout error. For that kind of situation, we can simply change some npm configurations like fetch-retriesfetch-retry-mintimeoutfetch-retry-maxtimeout and cache-min to minimize these timeout errors.

fetch-retries

This config controls the number of times npm try to connect to the registry when fetching packages. The default value is 2, but you can increase it to 3~5 or even more if you like. Open a terminal window and run the following command with the desired number of retries at the end.

npm config set fetch-retries 3

fetch-retry-mintimeout

This config controls the minimum time (in milliseconds) npm wait before timing out when fetching packages from the registry. The default value is 10000 milliseconds (10 seconds), change it to 100000 or more.

npm config set fetch-retry-mintimeout 100000

fetch-retry-maxtimeout

This config controls the maximum time (in milliseconds) npm wait before timing out when fetching packages from the registry. The default value is 10000 milliseconds (10 seconds), change it to 600000 or more.

npm config set fetch-retry-maxtimeout 600000

cache-min

This config controls the minimum time (in seconds) to keep items before re-checking the registry. The default value is 10 seconds, change it to 3600 seconds (1 hour) or more.

npm config set cache-min 3600

Finally, run this command to check whether all the configuration changes are successfully applied or not.

npm config ls -l

Now try to run some failing commands to check whether they are working. The duration values mentioned in this question was only for demonstration, you may try different values by increasing and decreasing them according to your situation. Visit npm-config documentation for more available configurations.

Using your GPU to make a lava lamp




We can use the gpus and some trignometric functions to make a digital lava lamp with the GPU. If you only see a black square, try a different browser that supports WebGL and be sure to turn on the WebGL support for any browser.
Iterations
Frames Per Second

 

Using your GPU to display the Mandlebrot Set




This Mandlebrot set is being displayed in your browser using your GPU. This is not a movie. If you only see a black square, try a different browser that supports WebGL and be sure to turn on the WebGL support for your browser. If only Benoit Mandlebrot, an IBMer, had a GPU on which to run his JavaScript code…
If not showing, try this in the chrome browser. link in the chrome browser.
Mandelbrot Iterations
Frames Per Second

 

After IBM Watson Trend, Watson has been put to work for Marketing Practitioners

Our IBM Incubator team built the next big thing for Watson.

This video really shows how new technology will assist practioners to work more in depth with their information and accomplish more than they can today. The practioners will be able to tackle problems with alacrity that are just not being tackled today.

I do not see this type of technology as replacing humans with computers. Humans don’t currently have the capability as the amount of data and processing required is already too much for current staffing levels to handle. What I see happening is that businesses that adopt and use this technology will grow and have the resources to bring in more people that generate the value unlocked in partnership with this technology.

Please, sit down and enjoy.

IBM Watson Trend wins 2016 DAA award for Excellence

Last night at the 2016 Digital Analytics Association Excellence Awards Gala, the application my team built, IBM Watson Trend, received the Technology of the Year Award! This is a notable industry award – especially given the competition, which included both top tier firms as well as innovative startups, such as Adobe and Ensighten. What a fun night!

https://www.digitalanalyticsassociation.org/awards2016_finalists

More Love and Press for the Vision of IBM Watson Trend

The author, Luke Bilton at InPublishing, had some great suggestions for future releases of our IBM Watson Trend application

An App I Like: IBM Watson Trend

Some excerpts from his article follow.


“IBM Watson Trend is an intriguing example of what automated content might look like in an app.

The app uses natural language processing and machine learning to intelligently analyse millions of conversations across social networks, blogs, forums, comments, ratings and reviews. These information sources are curated into a real-time buyers guide, forecasting the products which will be ‘hot’ and forecasting consumer interest.”

“It is a pleasing bundle of different content types which works well together as an editorial package.”

“It’s intriguing stuff. Now, if only I could connect Watson Trend (so I know what to buy) to my Google Calendar (so I don’t miss a birthday) and then to Amazon (so I don’t have to go to the shops)… I might automate my way to becoming the best dad in the world.”