News from the ITS Help Desk - September 2024

Help Desk Hours and Location

Just a quick reminder that the ITS Help Desk is here to support you. We are located in the Library on the Santa Maria campus and offer in-person support from Monday to Thursday, 8:00 AM to 4:30 PM, and Fridays from 8:00 AM to 4:00 PM.

Updates to ithelp.hancockcollege.edu

We will be implementing a new Something Broke ticketing form, targeting the beginning of October. A number of other services and features are also due for improvement. Updates will follow in future newsletters. Additionally, some changes will be made to the ticketing portal to attempt to improve readability. 

Essential Articles and Ticket Requests

Why Submitting a Ticket at ithelp.hancockcollege.edu is the Fastest Way to Get IT Support at Allan Hancock College

Imagine walking into a busy restaurant where customers shout their orders to the chef. Some scribble their orders on a napkin, some send emails, and others yell across the room. Chaos, right? It would be unreasonable to expect the kitchen staff to efficiently prepare meals with so many different and inconsistent methods of ordering. The same concept applies to IT issues at Allan Hancock College. When it comes to solving problems related to computers, software, access, and network concerns, the best way to get help is by submitting a ticket at ithelp.hancockcollege.edu. Let’s explore why this ticketing system is the most efficient way to get assistance.

1. Emailing a Technician Directly

Many people think emailing a specific technician might expedite their request, but in reality, it causes delays. Technicians spend the majority of their day out in the field addressing existing tickets in the order they were received. They’re not constantly checking personal emails, and if that technician is unavailable—whether due to sickness, vacation, or other assignments—your request could easily slip through the cracks. By submitting a ticket instead, your issue is guaranteed to be seen by the whole IT department, and anyone available can jump in to assist.

2. Contacting Senior Management First

Reaching out to senior management for assistance may seem like a good idea, but it actually slows down the process. Managers are typically not the ones resolving technical issues; they end up serving as middlemen, either submitting the ticket themselves or asking the help desk to do so. This adds unnecessary layers of communication and delays in getting your issue to the right person. Additionally, it can strain the system by pulling managers away from their own responsibilities.

3. Emailing helpdesk@hancockcollege.edu vs. ithelp@hancockcollege.edu

It’s important to know the difference between emailing helpdesk@hancockcollege.edu and ithelp@hancockcollege.edu. The first email address is designed for general questions, while the second automatically creates a ticket for technical issues. Sending an email to the general help desk for a problem that requires action may lead to delays or even missed requests. For the fastest resolution, always use ithelp@hancockcollege.edu, which directly generates a ticket for any technician to respond.

4. Asking a Technician in Passing

While it might feel convenient to mention your issue when you spot a technician nearby, this can cause problems. By doing so, you’re essentially cutting in line, bypassing others who submitted tickets and have been patiently waiting for a response. Moreover, the technician may not be prepared to address your issue at that moment, leading to inefficiencies. Always report your issue properly through the ticketing system to ensure fair treatment for everyone.

5. Calling the Help Desk

Calling the Help Desk is another option, but it's not the most time-efficient. When you call, the help desk staff still needs to listen to your explanation and then submit a ticket on your behalf. This adds an extra step and may delay the process. By submitting the ticket yourself, you provide all the necessary details upfront, making it quicker for the technician to start working on the issue. It is important to note that the Help Desk is available to assist with ticket creation if needed. 

6. Submitting a Ticket via ithelp@hancockcollege.edu or at ithelp.hancockcollege.edu

The most efficient and reliable way to report any IT issue is through the ticketing system at ithelp.hancockcollege.edu. This system ensures that the entire department is aware of your problem, and tickets are handled in the order they’re received. In urgent cases, such as when a classroom’s operations are completely disrupted, technicians can be redirected to prioritize emergency tickets. However, for everyday concerns, submitting a ticket is the fastest and most straightforward way to get help, ensuring your issue is resolved as quickly as possible.

Understanding AI: What It Is and What It Isn’t

Artificial Intelligence (AI) is a powerful tool that refers to machines or software capable of performing tasks that typically require human intelligence. This can range from simple functions, like recognizing speech, to more complex tasks like decision-making and problem-solving. AI systems, including machine learning and natural language processing, are designed to mimic certain human capabilities but are ultimately driven by data and algorithms.

However, it's essential to clarify what AI is not. AI is not conscious, self-aware, or capable of independent thought. It operates based on predefined rules and learned patterns from large datasets. While AI can simulate intelligent behavior, it doesn’t possess emotions, creativity, or a true understanding of the tasks it performs. It’s important to recognize these limitations, especially as the technology becomes more integrated into various industries.

Why AGI (Artificial General Intelligence) Doesn't Exist Yet

Artificial General Intelligence, or AGI, is often depicted in science fiction as a machine that can perform any intellectual task a human can. While today's AI systems are excellent at specific tasks—like translating languages or identifying patterns—they lack the broad, adaptable intelligence of a human mind. AGI remains a theoretical concept because current AI models cannot transfer knowledge between unrelated tasks, think creatively, or act autonomously across different domains. Achieving AGI would require advancements in understanding consciousness and intelligence, challenges that the AI field is still far from solving.

Can AI Be Detected? Risks of False Positives

Detecting AI-generated work in writing and coding is challenging. While detection tools, like Turnitin, scan for patterns or markers of AI involvement, they aren’t always reliable, especially when content is refined after generation. Human-edited AI output often becomes indistinguishable from original work, making detection more difficult.

False positives are a growing concern, as seen in the widely reported case of a Texas student accused of using AI for an essay (Student Wrongly Accused of AI Cheating By New Turnitin Detection Tool (rollingstone.com)). These mistakes occur when detection tools misinterpret formulaic or simple human writing as AI-generated, highlighting the limitations of these systems. Relying solely on detection software can lead to unjust penalties, so it's crucial to use these tools with caution.

For more insights into AI and education, check out ChatGPT and Beyond: How to Handle AI in Schools | Common Sense Education.

Key AI Vocabulary:

  1. Artificial Intelligence (AI): Machines or software designed to perform tasks that typically require human intelligence, such as recognizing speech, solving problems, or making decisions.
  2. Machine Learning (ML): A subset of AI that allows computers to learn and improve from experience without being explicitly programmed. It’s how AI systems get "smarter" over time by analyzing data.
  3. Natural Language Processing (NLP): A branch of AI focused on the interaction between computers and human language. It allows AI to understand, interpret, and generate human language, as seen in chatbots or translation tools.
  4. Generative AI: AI systems that create new content, such as text, images, or code. Tools like ChatGPT or DALL·E are examples, producing responses based on the prompts they receive.
  5. Prompt: The input or question given to an AI system to generate a response. For example, when you ask ChatGPT a question, that question is your "prompt."
  6. Model: A mathematical framework used by AI to recognize patterns and make predictions based on data. Models are trained using vast amounts of data to perform specific tasks, like writing or identifying images.
  7. Training Data: The information used to teach an AI model how to perform tasks. For example, a chatbot might be trained on millions of conversations to learn how to respond to different questions.
  8. Hallucinations: When an AI generates information that is incorrect or entirely fabricated, even though it may sound plausible. AI hallucinations occur because the model is designed to predict the next words in a sequence, not always to ensure factual accuracy.
  9. Overfitting: A situation where an AI model learns the training data too well, including noise or irrelevant details, which reduces its ability to generalize to new, unseen data.
  10. Underfitting: The opposite of overfitting, where a model is too simple and doesn't learn enough from the training data, resulting in poor performance on both training and new data.
  11. Bias: AI systems can inherit biases from the data they are trained on. If the training data is skewed or unbalanced, the AI may produce biased or unfair results.
  12. Token: A small unit of text (such as a word or part of a word) that AI systems process to generate responses. In tools like ChatGPT, prompts and responses are broken into tokens for efficient processing.
  13. Temperature: A setting that influences how creative or random an AI’s output will be. A higher temperature results in more creative and diverse responses, while a lower temperature produces more focused and predictable outputs.
  14. Epoch: One complete cycle through the entire training dataset during the training process of an AI model. Multiple epochs are often required to teach an AI model effectively.
  15. Fine-Tuning: A process where an AI model is further trained on a specific dataset to improve its performance in a particular task. For example, a general AI model might be fine-tuned to generate medical reports.
  16. Transformer Model: A type of neural network architecture that powers many modern AI systems, including ChatGPT. It allows AI to process and generate language more efficiently by focusing on important parts of the input.
  17. Inference: The process of making predictions or generating responses using a trained AI model. Every time you interact with an AI tool, it’s running an inference based on your prompt.
  18. Reinforcement Learning: A type of machine learning where an AI learns by trial and error, receiving rewards or penalties for actions it takes. This method helps improve performance over time.
  19. LLM (Large Language Model): A type of AI trained on vast amounts of text data, allowing it to generate coherent, contextually relevant responses. Examples include GPT (Generative Pre-trained Transformer) models like ChatGPT.
  20. Zero-Shot Learning: When an AI can perform a task it hasn’t been explicitly trained on, simply by applying general knowledge it gained from training on other tasks.
  21. Few-Shot Learning: An AI system's ability to perform a task after being provided only a few examples in the prompt. This shows how adaptable a model is to new tasks with minimal training data.
  22. Token Limit: The maximum number of tokens an AI system can process in a single prompt. Exceeding the limit can result in incomplete responses or errors.
  23. Context Window: The amount of text an AI model can consider at once when generating a response. For example, in a conversation, the context window includes everything the model "remembers" from earlier in the dialogue.
  24. Embedding: A way of representing text (or other data) in a format that AI models can understand. Embeddings transform words or phrases into numerical vectors, allowing AI to detect patterns and relationships between them.
  25. Deterministic: When an AI model generates the same output every time it is given the same prompt, without any randomness involved. This contrasts with probabilistic or random outputs, which vary slightly each time.

For a clear, visual explanation of current AI, check out IBM’s short video on the subject:

AI, Machine Learning, Deep Learning and Generative AI Explained (youtube.com)

This article was written with AI assistance.

Updates from Technology Council and Committees

  • Web Services Committee
    • Schedule: Meetings are held bi-monthly, usually on the 1st Thursday, from 9:30-11:00 am.
    • Update: The email widget is being addressed to improve functionality. Student digital ID cards, available on the portal, are being considered for improvement as a digital photo ID card. A tentative goal has been set to have these ready by Fall of 2025. 
  • Educational Technology Advisory Committee (EdTAC)
    • Schedule: Meetings are held bi-monthly, usually on the 1st and 3rd Wednesday, from 2:30-4:00 pm.
    • Update: The classroom standards document was discussed. Action has been taken to focus on improving laptop connections to podiums via USB-C connections. Classrooms that are up to date, in progress, or require in-depth overhauls toward the technology standard will be identified. 
  • Banner Committee
    • Schedule: Meetings are held monthly, usually on the 2nd Monday, from 9:00 - 10:30 am.
    • Update: November 15th we will begin migrating Banner hosting from the Santa Maria data center to Amazon Web Services. Banner and it's assorted resources will be unavailable from 12:00PM through Monday morning (Nov 18th).
  • Technology Council
    • Schedule: Meetings are held bi-monthly, usually on the 1st and 3rd Wednesday, from 2:30-4:00 pm.
    • Update: Set the following goals:

    1. Review progress in automating SEP processes (Banner committee) 

    2. Monitor effectiveness of Successnet widget and other portal widgets for notifying students of academic issues (Web Services committee) 

    3. Create Curriqunet websites with program map data (Web Services committee) 

    4. Upgrade existing podiums to allow for USB-C connections for data and video, on podiums that are ready for USB-C (edTac) 

    5. Create inventory of classroom technology (edTac) 

    6. Collect accreditation evidence 

    7. Partner with academic senate to create guidance on AI 

    8. Ensure that there is effective communication with stakeholders about significant technology changes (accreditation plan of action) 

August 2024 Ticket Summary

Here’s a brief overview of the Help Desk activity for July:

Total Tickets (Whole Department)

  • June: 453
  • July: 523
  • August: 682

Location-less Tickets:

  • Total Count: 420

Top Ticket Forms Without a Location:

  • Computer Help (for ithelp@hancockcollege.edu): 98 tickets
  • I Can't Sign In: 59 tickets
  • New Banner User or Banner Security Update: 45 tickets
  • Request Computers and Equipment: 29 tickets
  • In-Person Support Notes: 26 tickets

Lompoc Valley Campus (LVC) Tickets:

Total Tickets for LVC: 25

Top 5 LVC Buildings:

  • LVC 2: 8 tickets
  • LVC 5: 8 tickets
  • LVC 1: 8 tickets
  • LVC 3: 1 tickets

Top 5 LVC Ticket Forms:

  • Something Broke: 13 tickets
  • Install Something: 7 tickets
  • Audio/Visual Event Request: 2 tickets
  • Surplus Pick Up: 1 ticket
  • VOIP Telephone Request: 1 ticket

Santa Maria Valley Campus (SM) Tickets:

Total Tickets for SM: 237

Top 5 SM Locations:

  • Building A: 45 tickets
  • Building M: 31 tickets
  • Building C: 21 tickets
  • Building G: 20 tickets
  • Building O: 14 tickets

Top 5 SM Ticket Forms:

  • Something Broke: 130 tickets
  • Install Something: 45 tickets
  • Audio/Visual Event Request: 18 tickets
  • VOIP Telephone Request: 14 tickets
  • Classroom/Cart Maintenance Request: 10 tickets
Print Article