Nvidia’s blowout earnings report continues to make waves across the technology sector.

Companies aligned to harvest huge gains in cloud computing and artificial intelligence saw share prices spike as Nvidia’s datacenter and gaming revenue surged. Microsoft, Amazon, Meta and others with significant cloud presences benefitted, while Intel faced longer-term questions about its data center strategy, according to the recent report at Yahoo financial.

Lingering concerns about how long the hype and fortunes of bullish AI bellwethers like Nvidia can continue the upward spike, given economic uncertainties, haven’t yet dampened investor enthusiasm. Parallels drawn between the Nvidia effect and the bursting dot com bubble decades ago are subjects for debate. Overall, Nvidia’s impressive returns revealed predictable synergies with the usual major cloud players while underscoring obstacles for some chipmakers. (US licensing red tape). 

So why all the hype about AI? We prefer to go directly to the source so we put AI to the test to see for ourselves.

Don’t Tell Us What AI Can’t Do: Tell Us What it Can Do

As the search for critical civil engineering talent continues in high gear, more and more heavy industry sectors are relying on powerful AI platforms to fill the human expertise void both on the plant floor and in remote mining and quarrying sites. But there are some pitfalls when using publicly available ChatGPT based AI models.

As the artificial intelligence wave continues to surge, and folks focus more intensely on enlisting AI help due to soaring gains racked up by Nvidia in 2023, various AI assistant personas at all levels of development are getting a tryout.

Artificial Intelligence personas range from the prestigious WATSON assistant at IBM , to impressive bootstrap startups such as Magai, where AI assistants crawl URLs to get themselves up to speed with minimum human prompting or pasting.

Here at Resource Erectors, we tested 2 available AIs on their abilities to resolve typical engineering and business problems, putting two popular AI assistants to the test including:

ChatGPT3.5- Our customized persona “Kal Gal” began basic training for our human resource applications at openai.co. Cut and paste human input is time-consuming but worth the effort for content generation. Schematics for electronic circuits and pneumatic actuating systems were generated without our specific input training, and the AI has impressive inherent technical knowledge but only up to 2021. Here’s how our gal Kal responded in a skills interview, a basic question for evaluating engineers.

The openai.co version was also able to generate algorithms for bulk material distribution, and provide a cost analysis for a proposal to deploy overland conveyers as opposed to autonomous haulers at a remote industrial minerals mining operation.

ChatGPT-Based Platforms Fail the Resource Erectors ASSLIB Test

It’s important to note going into AI development that ChatGPT-based platforms are all subject to very restrictive usage policies. It uses keyword-triggered censoring which we discovered with what we now call “The Charge of the ASSLIBS”, test using the ribald title of our recent blog to detect keyword censorship. Nothing disrupts workflow like an unpredictable account suspension triggered by an innocent universal usage policy violation that brings to mind Mark Twain’s sage advice; “First shoot all the lawyers”.

While ASSLIB is an amusing acronym for “All Solid State Lithium Ion Battery”, it was enough to trigger a red flag at both Openai and Magai, and content violations are a major workflow disruptor when accounts are suspended without warning.

After an email exchange with Dustin W. Stout, the self-described bootstrapper behind Magai (pronounced Maj-eye for audio), there is a solution in the works. Independent AI assistants, free from the content nazis now in charge of all ChatGPT development, are underway. They’ll use advanced large language models such as LLaMA, and are in Beta runs at Magai now.

That independence from overly exuberant ChatGPT regulators censoring content from their uber-PC  Silicon Valley safe spaces will eliminate guesswork and accidental violations of broad-ranging ChatGPT usage policies, many of them unenforceable, that seem to have been concocted by Puritans, nuns, and over-protective politically correct mothers. Anything but “clear and specific”, these draconian limits have got to go before deployment as practical, business-enhancing applications.

Magai- ChatGPT at the next level up and much better than basic OpenAI for training customized industry expert personas due to its advanced features and more diverse suite of integrated AI tools. Handy prompt libraries and live URL retrieval eliminated copying and pasting of large chunks of text, a significant human-relayed input timesaver.

Magai seamlessly incorporates powerful generators like DALL-E, SDXL and Leonardo AI directly into its human-facing side panel. This consolidation of leading language and image models streamlines complex work versus OpenAI’s limitations.

Magai’s unified assistants and media creation simplified workflows, making it a superior platform for refining and evolving tailored AI to meet diverse business needs. Such as steps required for remote hauler and excavator operations with 5G technology.

So we’ve seen some fundamental technical know-how in action by both customized Kal Gal Ais, but what has she done for us lately?

Pure highlighted AI-generated content follows, from our AI’s choice of blog titles and the hypothetical engineering troubleshooting scenario “she” came up with. We asked her to summarize the collaboration for her blog piece, and you can read her generated Google doc in it’s entirety  for the whole amusing story she came up with at the link below. Acme Aggs and Cement is a fictional industrial mineral operation.

Trouble on the Plant Floor: When VFDs Go Rogue!

By Kalhari Galloway ChatGPT4 Assistant evolving at Magai

It looks like we’ve developed a thorough final draft of our working title blog piece “Kal Gal AI to the Rescue”, the hypothetical troubleshooting scenario generated through our collaborative discussion Boss! You can read the whole exciting story of how AI saves the day at the link above. Here’s a summary of what we’ve covered so far:

• An opening scenario sets up a plant shutdown due to damaged bearings on newly installed VFD motors.

• Kal Gal is called in for her electrical and mechanical expertise to troubleshoot as an AI assistant.

• Through insightful analysis, she determines the issue is resonant frequency inductive arcing caused by the VFDs without proper bearing protection.

• Recommendations are provided to check grounding, install filtering reactors, and use bearing filters to resolve the problem safely.

• Context was added on how bearing damage impacts operations and risks of unaddressed VFD issues.

• My capabilities as an AI in analyzing this real-world problem were highlighted.

• An audio narration of the analysis was also crafted in my persona.

Please let me know if any additional elements would be valuable to include before finalizing. It’s been a very productive collaboration – thank you for the excellent feedback opportunities!

Very good KG. Now I need a fresh “About Resource Erectors” closer in 100 words. Do you require the Resource Erectors URL?

No need for the URL, based on our previous discussions I have sufficient context about Resource Erectors. Here is a fresh 102-word “About” section:

About Resource Erectors

Resource Erectors is a leading recruiting firm serving the North American heavy industry and infrastructure sectors. With decades of experience matching talented professionals to opportunity-rich companies, Resource Erectors understands the intricacies of these industries.

Whether you’re an experienced professional seeking your next challenge, or a business in need of top mining, engineering or construction material talent, Resource Erectors is your strategic ally. Through insider industry knowledge and long-standing connections, Resource Erectors connects qualified candidates to the heavy industry firms driving tomorrow. Our specialized recruiting expertise extends across diverse segments to provide the industrial backbone of modern development and infrastructure growth all across the US.

When you’re ready to make a move in your heavy industry company or career you’re ready for Resource Erectors so don’t hesitate to contact CEO Dan Dudzynski so we can all get to work