KNOWLEDGE AND INTELLIGENCE
Karl Dakin
Capital Coach | Stakeholder Investor Campaigns | Design, Stage, and Manage or Support | Reduce Time, Money, and Risk of Raising Funding | Increase Probability of Success! | Opportunity Management
CONTENTS
?·?????? Quick Calendar
·?????? Funding Point
·?????? 1 Year Anniversary – Giving to Kiva
·?????? Capital Coach
·?????? Don Cohen Show
·?????? Subscribe – Daily or Weekly Editions
?
?QUICK CALENDAR
?I quickly list here upcoming opportunities where I will be sharing information, lessons in funding, or meeting in person or via videoconferencing.
?June 12 - SuperCrowd Chicago – Workshop on Crafting Crowdfunding Offers
?
FUNDING POINTS – KNOWLEDGE AND INTELLIGENCE
?As I learn of new AI tools, work to master use of selected AI tools, and continue my joint venture with Knowledge Avatars to advance CAP – my own AI avatar, I see the familiar challenges of adopting innovative technology: misuse, misunderstanding, fear, and governments seeking to implement new regulations.
?Emiliano De Laurentiis, the CEO and Founder of Knowledge Avatars, was a guest on my Successful Funding show earlier today where we talked about key questions about what is called artificial intelligence (AI). We demonstrated CAP by giving it prompts that generated responses out of proprietary data that I had provided and generated responses out of OpenAI’s Chat GPT 4.0.
?The key challenge is distinguishing between knowledge and intelligence.
?People are using Large Language Models (LLMs) like Chat GPT or Gemini.
?Large language models?are computer programs capable of generating human language by processing vast amounts of text data.
?I take the position that large language models are not intelligent. Giving them the adjective of ‘artificial’ does not clarify this issue.
?I intend to take a sidestep in achieving a definition of artificial intelligence and come at this issue from a different angle: Can a person or artificial intelligence answer your question correctly?
?If you ask a question of a large language model (make prompt), you will get a response that looks like an answer to your question. But it is not. What the large language model did was assign a value to your prompt and construct a response that represents an average of all the answers it has seen given to associated questions to which it has access. This is not an answer. It is data.
?If you give a prompt that matches a large volume of usual questions for which there is a relative consensus in providing an answer, you will receive data that looks like an answer to a question that may provide you with the information you are seeking.
?If you give a prompt that does not match a large volume of usual questions or for which there is no relative consensus in providing an answer, you will receive data that looks like an answer to a question that may be correct or incorrect. If you cannot tell if the response is correct or incorrect, then your use of the response may put you or your small business at risk.
?If you give a prompt that cannot be matched with usual questions or for which there is no public information, you will receive data that looks like an answer to a question that is probably incorrect. The large language model has crafted a response that is not based on fact.
?A person is left with the following choices:
·?????? Do not use large language models (labeled as artificial intelligence or otherwise)
·?????? Use large language models carefully
§? A common body of questions
§? A consensus of answers to questions
?In developing CAP with Knowledge Avatars, we are working to modify large language models using learning management systems (LMS) and proprietary knowledge. Training so far has been limited to the topic of ‘funding’ within the large context of entrepreneurship and small business operations. As a result, prompts to CAP result in better responses on the funding of small businesses. The responses are still not answers that are paired precisely with questions regarding the unique situation of a person or small business seeking information. However, it may be close enough that a person may depend upon it. It may not.
?If a person comes to me and poses a question (not a prompt), I can give them:
·?????? An answer,
·?????? An opinion, or
·?????? An “I don’t know”
In each of these answers, I am responding to a question in the context of the person asking the question. I am not providing an average of a group of answers to a group of questions that may or may not have anything to do with the person asking the question.
领英推荐
?When using a large language model, a person must recognize they are working with a computer system that is averaging data and they are not working with a sentient being. No matter how closely a response to a prompt may appear to be an answer, it is just data.
?Clearly, large language models represent a new tool with more capabilities which may offer all kinds of benefits.
?I recommend the use of large language models with caution. You need to ask yourself whether you are getting knowledge or data, no matter how similar they may appear. If you cannot tell the difference, implement appropriate action to protect yourself against the possibility that the response is incorrect. This recommendation is true for the use of experts and consultants. When you ask for advice, you may get different recommendations that may prove correct, incorrect, or somewhere in between.
?Where possible, determine where you can enhance your use of large language models by:
·?????? Verifying responses
·?????? Combining large language models with other tools
·?????? Modifying large language models to provide better responses
?Keep in mind that large language models or products and services that incorporate large language models are tools. Masterly requires practice and training.
??
1-YEAR ANNIVERSARY – GIVING TO KIVA
I am celebrating my first anniversary in giving to Kiva, resulting in no-interest loans to businesses worldwide.
CAPITAL COACH
?As seen on today's Successful Funding show.
I am now field evaluating the Capital Coach, an artificial intelligence learning management system avatar (AILMSA) that I have trained on the topic of funding. The Capital Coach is embedded as a widget on my Dakin Capital website at https://dakincapital.com/CAP. The Capital Coach represents a proprietary data set on the Knowledge Avatars learning management system with licensed access to Open AI. Could you give the Capital Coach a try and ask it questions about funding?
I am working with Knowledge Avatars to make CAP the ultimate tool to support small businesses and community projects in raising funding. Two current challenges are (1) training the AI avatar to match specific funding situations and (2) explaining that a highly personalized AI avatar is so much more than a chatbot powered by a Large Language platform. Unique, non-public information about an individual or an organization cannot be obtained through a chatbot. The recent advances in AI have narrowed the field of prospects who stand to gain from personalized AI to high-net-worth individuals, professional service providers, organizational leaders, and other similarly placed who work with fast-moving, fast-changing information that may never be public.
Please share your feedback. Karl Dakin at [email protected]
??
DON COHEN SHOW
Yesterday, I was a guest on Don Cohen’s show where we talked about promotion – the act of sharing information to gain awareness and solicit action. You may see a recording of the show at:
I enjoy being a guest, engaging in conversations with other guests, and building communities with Don, using LinkedIn as a communication channel. Every show is its own adventure as we start with a word of the day and then see where the conversation goes from there.
Don is an expert on LinkedIn, particularly on the use of Live streaming to build brands and communities. We will discuss using LinkedIn as a social media platform for building communities that support raising funding.
All shows where I have been a guest can be viewed on Don Cohen's LinkedIn page under Posts.
??
SUBSCRIBE - THE DAILY EDITION OR THE WEEKLY EDITION
Please subscribe to and share this newsletter.
This Instant Funding newsletter is now available in a Daily Edition or a Weekly Edition. You may choose between this Edition, published seven times a week, or the Weekly Edition, published once a week. If you are already subscribed to the Daily Edition and wish to change, just ‘unsubscribe’ and then ‘subscribe’ to the Weekly Edition.
?
?Karl Dakin, the Capital Coach
Dakin Capital LLC
?
?