YT YouTube2Text

YouTube2Text → Transcripts → PROJECT LUMIÈRE TASKING PROCESS (CHEMISTRY)| YOU NEED ANSWERS? TEXT ME WHATSAPP +254700123813

PROJECT LUMIÈRE TASKING PROCESS (CHEMISTRY)| YOU NEED ANSWERS? TEXT ME WHATSAPP +254700123813 — Transcript

by Amazing Tv 254 · 2,866 words · 491 segments · language en · Watch on YouTube

Open in transcriber Summarize with AI Download SRT/VTT

Full transcript

0:00Hello everyone, welcome back to my
0:02channel. Today I want to talk about
0:04project Lumiere, the chemistry project
0:06on Handshake AI. As you can see on the
0:08screen, project Lumiere is listed under
0:11the current project section. The project
0:13is remote and offered as a contract
0:15opportunity. It currently shows a
0:17payment of $110 per task, which makes it
0:20an interesting opportunity for people
0:22with a background in chemistry and other
0:24STEM related fields. In this video, I
0:27want to take some time to explain what
0:29this project is about and what you
0:30should understand before moving forward.
0:33We're going to look at the type of work
0:34involved, the importance of image-based
0:37questions, the assessment process, and
0:39some of the things you should pay
0:40attention to when working through the
0:42project. Project Lumiere focuses on
0:44creating challenging image-based
0:47questions in the field of chemistry. The
0:49project is designed for STEM
0:51professionals who have a strong
0:52understanding of scientific concepts and
0:55who can create questions that require
0:56reasoning and careful analysis. This is
0:59an important point because the purpose
1:01is not simply to create ordinary
1:03chemistry questions. The questions are
1:05intended to challenge an AI system. The
1:08AI needs to understand the information
1:10presented, interpret the visual
1:12material, follow the instructions, apply
1:14relevant chemistry knowledge, and
1:16eventually arrive at an appropriate
1:18answer. This means the person creating
1:20the question needs to think carefully
1:22about the entire problem. You are not
1:24simply writing a question and attaching
1:26an image. You need to consider how the
1:29image supports the question and whether
1:31the information presented allows the AI
1:33to reason through the problem. The
1:35imagebased component is one of the most
1:38interesting parts of project Lumiere.
1:40Chemistry is a subject where visual
1:42information can be extremely important.
1:44Chemical structures, molecular diagrams,
1:47reaction pathways, laboratory setups,
1:50graphs, tables, equations, and other
1:53scientific illustrations can communicate
1:55information that may not be easy to
1:57describe using words alone. For example,
2:00a question might include a chemical
2:02structure and ask the model to identify
2:04a particular functional group. Another
2:06question could show a chemical reaction
2:08and ask what product would be expected
2:10under specific conditions. A task could
2:13also provide a graph showing
2:14experimental results and ask the model
2:17to interpret the relationship between
2:19different variables. There could also be
2:21questions involving molecular
2:22structures, stereochemistry, reaction
2:25mechanisms, equilibrium, thermodynamics,
2:28acids and bases or other chemistry
2:30concepts. The exact content will depend
2:33on the requirements of the project. The
2:35important thing is that the image should
2:37have a clear purpose. The visual should
2:39provide information that is relevant to
2:41solving the question. It should not
2:43simply be added because the project
2:45requires an image. This distinction is
2:47very important when creating highquality
2:50tasks. A strong image-based question
2:52should encourage the AI to examine the
2:55visual information carefully. The model
2:57should not be able to ignore the image
2:59and answer the question immediately from
3:02general knowledge. If the question can
3:04be answered without looking at the
3:05image, then the image may not be
3:07contributing enough to the task. This is
3:10why contributors need to think about the
3:12relationship between the image and the
3:14question. The question should encourage
3:16the model to use both sources of
3:18information. The written instructions
3:20provide the problem while the image
3:22provides important evidence or
3:24information needed to solve that
3:26problem. When these two elements work
3:28together, the task can become much more
3:30useful for evaluating AI reasoning. When
3:33you open the project, you can use the
3:35continue button to move through the
3:36required steps. In this example, the
3:39dashboard shows that there are five
3:40steps left before the project process is
3:43completed. These steps are important
3:45because they can help you understand
3:46what the project expects from
3:48contributors. Before beginning any
3:50assessment, I strongly recommend reading
3:53the instructions carefully. Do not
3:55immediately click through the project
3:56without understanding what you are being
3:58asked to do. Projects on Handshake AI
4:01can have different requirements. One
4:03project may involve evaluating images.
4:06Another project may involve writing
4:08questions. Another may focus on coding,
4:10reasoning, language, science, or other
4:13specialized areas. Therefore, you should
4:16always focus on the specific
4:17instructions provided for the project
4:19you are currently working on. For
4:21project Lumiere, having chemistry
4:23knowledge is obviously important.
4:25However, scientific knowledge is only
4:28one part of completing this type of
4:29work. You also need to understand how to
4:32communicate scientific information
4:34clearly. A chemistry question can be
4:36scientifically correct but still be
4:38poorly designed. For example, the
4:40wording could be confusing. The image
4:42could contain unnecessary information.
4:45The question could have multiple
4:47possible interpretations. The expected
4:49answer could be unclear. Or the image
4:52may not contain enough information to
4:53solve the problem. All of these factors
4:56can affect the quality of the task. That
4:58is why careful review is so important.
5:01When creating a question, start by
5:03thinking about what you want the AI to
5:04determine. Then consider what
5:06information the AI needs to reach that
5:08conclusion. After that, you can design
5:11the image and question around that
5:12reasoning process. This approach can
5:15help create a more purposeful task. The
5:17question should also have an appropriate
5:19level of difficulty. A question that is
5:21extremely easy may not provide enough
5:24challenge. At the same time, a question
5:26that is unnecessarily complicated may
5:29become confusing. The goal is to create
5:31a meaningful challenge. The AI should
5:34need to analyze the information and
5:36apply its knowledge. This could involve
5:38several steps of reasoning rather than
5:40simply recalling a fact. For example,
5:43instead of asking the model to identify
5:45a chemical term from memory, the task
5:47could provide a structure, reaction, or
5:50experimental result and ask the model to
5:52interpret it. This makes the task more
5:54dependent on reasoning. Another
5:56important consideration is scientific
5:58accuracy. Chemistry requires a high
6:01level of precision. Chemical formulas
6:03need to be correct. Chemical structures
6:05need to represent the intended
6:07compounds. Equations should be balanced
6:10when appropriate. Numbers and units
6:12should also be checked carefully. The
6:14same applies to graphs and tables. If
6:16the image contains incorrect
6:18information, the question may test the
6:20AI using information that is
6:22scientifically inaccurate. This can
6:25reduce the usefulness of the evaluation.
6:28Therefore, contributors should carefully
6:30review every scientific detail before
6:32submitting a task. It is also important
6:35to check the readability of the image.
6:37If the image contains small labels,
6:39chemical structures, equations, or
6:42numbers, they need to be visible. An
6:44image can be technically correct but
6:46still difficult to understand if the
6:48important details are too small or
6:50unclear. Remember that the AI is
6:53expected to interpret the image. At the
6:55same time, you should not make the image
6:58unnecessarily complicated. Too much
7:00information can distract from the actual
7:02problem. A strong image should contain
7:04the information needed for the question
7:06without adding unnecessary details. This
7:09is where good task design becomes
7:11important. The person creating the task
7:13needs to decide what information is
7:15essential and what information can be
7:17removed. This requires careful thinking
7:20about the intended reasoning process.
7:22Another important area is the wording of
7:24the question. Try to make the question
7:26specific. The person answering should
7:29understand exactly what they are being
7:31asked to determine. Avoid unnecessary
7:33wording that could introduce confusion.
7:35The question should also match the
7:37information shown in the image. If the
7:39question asks about one chemical
7:41structure, the image should clearly show
7:44that structure. If the question asks the
7:46model to compare two compounds, both
7:49compounds should be presented clearly.
7:51If the question asks about experimental
7:53results, the relevant data should be
7:55available in the image. The connection
7:58between the image and the question
8:00should be easy to understand. Another
8:02thing to consider is whether the answer
8:04can be supported by the information
8:06provided. The question should not depend
8:08on information that is completely
8:10missing. If a calculation requires a
8:12particular value, that value should
8:14either be provided or be appropriately
8:16available based on the project
8:18instructions. This is especially
8:20important for numerical chemistry
8:22problems. For example, if a question
8:24involves calculating a concentration,
8:27the necessary values and units should be
8:29clear. If a question involves
8:31interpreting a graph, the axes and
8:33labels should be understandable. If a
8:36question involves a chemical reaction,
8:38the relevant reactants and conditions
8:40should be identifiable. These small
8:43details can make a major difference.
8:45Project Lumiere may also involve
8:47different areas of chemistry. Chemistry
8:50is a very broad field. It includes
8:52general chemistry, organic chemistry,
8:54inorganic chemistry, analytical
8:56chemistry, physical chemistry,
8:58biochemistry, and many other specialized
9:01areas. Different questions may require
9:03different types of knowledge. Some
9:05questions may focus on basic concepts.
9:08Other questions may require more
9:10advanced scientific reasoning. Depending
9:12on the project guidelines, contributors
9:14may need to work with different levels
9:16of complexity. This is why it is
9:18important not to assume that every
9:20question should follow the same format.
9:23Instead, follow the instructions
9:24provided for the specific task. The
9:27assessment stage is another important
9:29part of the process. If you are new to
9:31Handshake AI projects, you may notice
9:33that some projects require you to
9:35complete an assessment before gaining
9:37access to certain tasks. The assessment
9:40can help determine whether you
9:41understand the project requirements. For
9:43this reason, it is important to take
9:45your time, read every question
9:47carefully, look at the examples
9:49provided, pay attention to the
9:51instructions, and most importantly, try
9:54to understand why a particular answer is
9:56considered appropriate. Do not simply
9:58choose an answer because it looks
9:59familiar. Some assessments may include
10:02details that are easy to overlook. A
10:04single instruction can change the way
10:06you should approach the question. This
10:08is why careful reading is so important.
10:10You should also remember that assessment
10:12questions may be designed to test your
10:14ability to follow instructions. Even if
10:17you have strong chemistry knowledge, you
10:19still need to follow the specific
10:20guidelines of the project. The project
10:23may define what makes a good question,
10:25what type of image is acceptable, how
10:27answers should be written, and how the
10:29final work should be evaluated.
10:31Following these requirements is part of
10:33doing the task correctly. Another
10:35important point is that quality should
10:37come before speed. When you see a
10:39payment amount such as $110 per task, it
10:42can be tempting to focus on completing
10:43tasks quickly. However, rushing can
10:46increase the possibility of mistakes. A
10:49better approach is to understand the
10:51requirements first. Once you become
10:53familiar with the process, you may
10:55naturally become faster, but speed
10:57should not come at the expense of
10:59quality. Take time to review your work
11:01before submission. One useful approach
11:03is to perform a final quality check.
11:06First, look at the image and make sure
11:08it is clear. Then read the question from
11:10beginning to end. Next, check whether
11:12the question actually depends on the
11:14image. After that, verify the scientific
11:17information. Finally, check the expected
11:20answer and make sure it follows the
11:22project instructions. This final review
11:24can help catch mistakes that may have
11:26been missed during the initial creation
11:27process. Another useful habit is to look
11:30at the task from the perspective of the
11:32AI. Ask yourself what information the
11:34model can see. Ask whether the important
11:36details are visible. Ask whether the
11:39question gives enough information to
11:41reach the expected conclusion. Always
11:43check your own Handshake AI account for
11:45the latest project information. Your
11:47dashboard may show different information
11:49depending on your eligibility and the
11:51availability of the project. If project
11:54Lumiere appears in your account,
11:56carefully review the requirements before
11:58deciding whether to continue. If you
12:00qualify and decide to participate, make
12:02sure you complete each required step. Do
12:05not skip the assessment instructions. Do
12:07not ignore project examples and do not
12:10assume that another project follows the
12:12same rules. Every project can have its
12:14own workflow. For new contributors, it
12:17may take some time to become comfortable
12:19with the platform. That is completely
12:21normal. The more projects you work
12:23through, the more familiar you can
12:24become with the general process of
12:26reading instructions, completing
12:28assessments, reviewing requirements, and
12:31submitting work. However, each new
12:33project should still be treated
12:35independently. Always start by
12:37understanding the specific requirements.
12:39Another thing I want to emphasize is the
12:41importance of consistency. If you create
12:44several questions for a project, try to
12:46maintain the same level of quality
12:48across them. Each question should meet
12:50the project requirements. Each image
12:52should be relevant. Each answer should
12:54be accurate and each task should be
12:57reviewed before submission. Consistency
12:59can be just as important as creating one
13:02excellent example. You should also be
13:04prepared to learn from feedback. If a
13:06project provides feedback on your work,
13:08use that information to improve future
13:10tasks. Maybe the question was too easy.
13:14Maybe the image needed more clarity.
13:16Maybe the wording was ambiguous. Maybe
13:18the answer explanation needed additional
13:20reasoning. Whatever the feedback may be,
13:23understanding it can help you improve
13:25your future work. This is particularly
13:27useful for contributors who plan to work
13:29on similar projects over time. As you
13:32gain experience, you may become better
13:34at recognizing what makes a strong task.
13:36You may also become more comfortable
13:38with different chemistry topics and
13:40different types of visual information.
13:42That experience can help you approach
13:44future assessments with greater
13:46confidence. For anyone interested in
13:48STEM, chemistry, education, scientific
13:51research, or artificial intelligence,
13:54project Lumiere is certainly an
13:55interesting project to learn about. It
13:58brings together scientific expertise and
14:00AI evaluation in a practical way. The
14:03project also shows how specialized human
14:06knowledge can play a role in the
14:07development of AI systems. Even as AI
14:10continues to improve, human expertise
14:13remains important. Experts can help
14:15create challenging examples, identify
14:17mistakes, evaluate responses, and
14:20provide information that can help
14:21improve the quality of AI systems.
14:24Projects like Lumiere are part of this
14:26broader process. The combination of
14:28chemistry and artificial intelligence is
14:31also becoming increasingly relevant. AI
14:33is being used across scientific
14:35research, education, drug discovery,
14:38data analysis, and many other areas.
14:41Because of this, understanding how AI
14:43handles scientific information can be
14:45valuable for people working in STEM
14:47fields. That makes projects like this
14:49interesting beyond the individual task
14:52itself. They provide an opportunity to
14:54see how scientific knowledge and AI
14:56technologies can work together. Before
14:59finishing, I also want to remind
15:01everyone to be careful with project
15:03information. If you see a project
15:05advertised with a specific payment
15:06amount, always verify that information
15:09inside your own account. Do not assume
15:11that every person will see exactly the
15:13same payment or availability. The
15:16information can depend on the project
15:18account eligibility and current platform
15:20conditions. Also, make sure you follow
15:23the official instructions provided by
15:25the platform. Do not rely only on
15:27information from videos, social media
15:29posts, or other contributors. Those
15:32sources can be helpful for understanding
15:34the general process, but the official
15:36project instructions should always be
15:38your main reference. So, if project
15:41Lumiere is available in your Handshake
15:43AI account, take some time to explore
15:45it. Read the project description. Review
15:47the requirements. Look through the
15:49assessment steps. Understand the type of
15:52chemistry questions being requested. Pay
15:54attention to the image requirements. And
15:57make sure you understand how the final
15:58task should be completed before moving
16:00forward. In this video, I will continue
16:03showing you what project Lumiere looks
16:05like on the Handshake AI platform. We
16:07will look at the project information,
16:09the available steps, and the different
16:11areas you should pay attention to before
16:13starting. I will also continue
16:15explaining how image-based chemistry
16:17questions can be approached and what
16:19makes these tasks different from
16:20ordinary chemistry questions. The goal
16:23is to give you a better understanding of
16:25the project and help you know what to
16:27expect when reviewing the available
16:28information. If you are interested in
16:30Handshake AI projects, online AI work,
16:33STEM opportunities, chemistry projects,
16:36assessments, and different ways to
16:38participate in AI training, make sure
16:40you subscribe to the channel. I will
16:42continue sharing videos about different
16:44projects, assessments, guidelines, and
16:47online AI opportunities. If you find
16:49this information useful, make sure you
16:51like the video and share it with others
16:53who may be interested in this type of
16:55work. You can also leave a comment and
16:57let me know which Handshake AI project
16:59you would like me to cover next. There
17:01are many different projects appearing on
17:03the platform and each one can have its
17:05own requirements and workflow. That is
17:07why I will continue breaking them down
17:09and explaining what you can expect from
17:11each project. Remember, always read the
17:14official instructions carefully. Take
17:16your time during assessments. Check your
17:18work before submitting and always verify
17:21the latest project information inside
17:23your own handshake AI account. Thank you
17:26so much for watching this video. I hope
17:28this extended explanation has given you
17:30a much better understanding of project
17:32Lumiere and the chemistry project on
17:34Handshake AI. If you are interested in
17:36learning more about this project, stay
17:38with me as we continue exploring the
17:41assessment and the different steps
17:42involved. Make sure you subscribe to the
17:45channel, turn on the notification bell,
17:47and stay connected for more videos.
17:49Thank you for watching, and I will see
17:51you in the next video.

About this transcript

This page contains the full transcript of PROJECT LUMIÈRE TASKING PROCESS (CHEMISTRY)| YOU NEED ANSWERS? TEXT ME WHATSAPP +254700123813 by Amazing Tv 254, generated from the public captions YouTube serves with the video. The transcript has 2,866 words across 491 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.

What you can do with it

Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to translate the transcript into 100+ languages.

Free YouTube transcript tool

YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, translation to 100+ languages, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.