Microsoft เกาะกระแสไวมาก พอเห็น OpenAI เปิดตัว ChatGPT มาแล้วมีผู้ใช้งานถึง 100 ล้านคนไวมาก กระแส Generative AI มาแรง Microsoft รีบเปิดตัวเลยว่า ได้เตรียมเอา ChatGPT ไปผนวกกับ Bing ที่เป็น Search engine ของตัวเอง เล่นเอา Google นั่งไม่ติด ต้องรีบออกมาบอกเหมือนกันว่า เฮ้ย! ฉันก็มีพัฒนา Generative AI เอาไว้นะ มันชื่อว่า Bard วันนี้ผมได้ลองมาครบทั้งสองตัวแล้ว เลยขอมาเขียนแปะเอาไว้หน่อยว่าผมสังเกตเห็นอะไรบ้าง
เอาจริงๆ ผมก็เชื่อว่าทั้งสองค่ายอะเตรียมทำกันมานานแล้วแหละ เพราะของพวกนี้มันไม่ใช่ปุบปับนึกจะทำก็ทำได้ การเทรน AI ให้เก่งมันต้องใช้ทรัพยากรและเวลามากมายครับ เพียงแต่เกมนี้ Microsoft ชิงเปิดได้ก่อนนั่นเอง
Search enginer + Generative AI = การเปลี่ยนแปลงของระบบเสิร์ช
ใครที่ติดตามเรื่อง ChatGPT มาก็น่าจะพอนึกออกว่าสิ่งที่มันทำได้มีอะไรบ้าง มันสามารถรับคำสั่ง (ที่เรียกว่า Prompt) จากเรา ซึ่งสามารถถูกเขียนในรูปแบบภาษามนุษย์ทั่วๆ ไปได้เลย ตัว AI มันจะถูกฝึกฝนมาให้สามารถถอดใจความสิ่งที่เราป้อนเข้าไปได้ จากนั้นมันก็จะเอาไปประมวลผล แล้วแสดงเอาต์พุต (Output) ออกมา โดยอ้างอิงจากข้อมูลใดๆ ที่มันถูกเทรนมานั่นเอง ซึ่ง ChatGPT เนี่ยถูกเทรนมาหลายเรื่องมาก แต่มีข้อจำกัดตรงที่ข้อมูลที่ถูกใช้เทรนมันจบที่เดือนกันยายน 2564

การที่ Microsoft เอา ChatGPT ไปผสมผสานกับ Bing เลยเป็นการทลายข้อจำกัดด้านข้อมูลออกไป เพราะนั่นเท่ากับว่า ChatGPT จะสามารถเข้าถึงข้อมูลจากอินเทอร์เน็ตได้นั่นเอง หลักการทำงานของมันคร่าวๆ ก็คือ ChatGPT จะทำการวิเคราะห์ว่าเราต้องการค้นหาอะไร ซึ่งเราจะสามารถพิมพ์ถามได้เหมือนการพูดคุยกับคนตามปกติเลย จากนั้นมันก็จะทำการค้นหาด้วยคีย์เวิร์ดที่มันวิเคราะห์ออกมาได้ และเมื่อได้ข้อมูลกลับมา มันก็จะทำการวิเคราะห์ข้อมูล แล้วนำเสนอข้อมูลนั้นกลับมาให้เราในรูปแบบที่พร้อมไปใช้งานมากที่สุด

ดูจากตัวอย่างการใช้งาน Bing ด้านบน จะเห็นว่าตอนผมถามไปว่า Who won the best actor from the latest oscar? แบบที่คนทั่วไปจะถามๆ กันเนี่ย ChatGPT ก็จะทำการวิเคราะห์ประโยคของผม แล้วพบว่าคีย์เวิร์ดสำหรับการค้นหาก็คือ best actor oscar winner latest และเมื่อค้นข้อมูลมาได้ ก็วิเคราะห์แล้วนำมาสร้างความเป็นคำตอบให้ผมว่า Brendan Fraser won the Best Lead Actor award at the 95th Academy Awards ceremony for his role in ‘The Whale’. Congratulations to him! แทนที่จะเป็นการบอกว่าให้เราไปอ่านข้อมูลเอาเองที่เว็บไหน
สิ่งนี้คือสิ่งที่เขามองกันว่ามันจะมาเปลี่ยนแปลงพฤติกรรมการเสิร์ชข้อมูลของคน เพราะว่าคนจะไม่จำเป็นต้องเข้าเว็บไซต์อีกแล้ว ในกรณีที่ต้องการหาข้อมูลแบบนี้ เพราะ Generative AI จะทำการอ่านเนื้อหาเว็บมาให้ แล้วเอามาสรุปเป็นคำตอบให้เราแบบตรงประเด็นเลย ซึ่งจริงๆ อะไรแบบนี้เป็นสิ่งที่ Google พยายามทำมานานแล้วนะ เพียงแต่มันไม่ได้ออกมาในรูปแบบแชทเหมือน Bing โฉมใหม่ที่รวมร่างกับ ChatGPT เท่านั้นเอง

แต่ความสามารถของ Generative AI มันมากกว่านั้นครับ เพราะเราอาจจะสามารถให้มันช่วยจัดหมวดหมู่ของผมการค้นหามาเป็นตารางก็ได้ หรือ ขอให้สรุปข้อมูลการค้นหามาในรูปแบบที่เราต้องการก็ได้ เช่น ตัวอย่างด้านล่างที่ผมถามว่าใครชนะรางวัลอะไรในงานออสการ์บ้าง แต่ให้นำเสนอในรูปแบบตาราง ทำให้อ่านดูง่ายขึ้น เป็นต้น

Microsoft Bing + ChatGPT vs Google + Bard
โดยความเห็นส่วนตัวของผม Microsoft นำก้าวแรกไปด้วยการเปิดตัว ChatGPT ซึ่งมันว้าวมาก แต่ถ้า Bard ก็ทำได้ว้าวไม่แพ้กันละก็ Microsoft ก็จะแพ้ในเกมถัดไปครับ เพราะสุดท้ายแล้ว แม้ Bing จะพัฒนาขึ้นไปเยอะมาก ทั้งในแง่ของความเร็ว และความแม่นยำของการค้นหา แต่ถ้าเทียบกับ Google แล้ว ก็ต้องบอกว่าเป็นรองอยู่หลายขุมมาก โดยเฉพาะในกรณีที่เราต้องการค้นหาข้อมูลแบบค่อนข้างเฉพาะเจาะจงหน่อย และเป็นภาษาท้องถิ่น (Bing แสดงผลการค้นหาภาษาไทยแย่กว่า Google เยอะมาก)
ดังนั้น หาก Google สามารถทำให้ Bard เจ๋งได้พอๆ กับ ChatGPT ละก็ Microsoft ก็จะแพ้ในเกมโดยรวมทันที เพราะมันจะไปแข่งกันที่ความสามารถของ Search engine แล้ว ว่าใครจะสามารถค้นหาข้อมูลมาได้ถูกใจผู้ใช้งานมากกว่ากัน
งั้นเรามาลองค้นหาโน่นนี่นั่นกันหน่อย ว่าเทียบระหว่าง Microsoft Bing + ChatGPT vs Google + Bard แล้ว เป็นยังไงบ้าง เริ่มจากคำถามแรกก่อนเลยครับ How generative AI will impact sear (สำหรับผู้พิการทางสายตา ผมได้เอาผลการค้นหามาใส่ในตารางด้านใต้รูปให้ด้วยนะครับ เพราะคงใส่เป็น Alternative text ไม่ไหว มันค่อนข้างยาวมาก)


Microsoft Bing | Google Bard |
---|---|
Generative AI is already changing search engines. It can create new content such as images, text, audio, video, and even code usually from a prompt or command. This means that generative AI-based search engines can become a problem for online research. As soon as these companies start planning revenue generation by getting payments to promote content from certain websites or publishers, the online search will become highly biased. The influence of artificial intelligence-generated content on search engines such as Google is a complex subject. On the one hand, AI-generated content can save content creators time and effort, but on the other hand, it may not always deliver high-quality and original content for the readers. In addition to this, integrating large language models into search engines could mean a fivefold increase in computing power and huge carbon emissions. | Generative AI is a technology that can create content, such as text, images, and videos. It is still under development, but it has the potential to revolutionize search engines. Generative AI could be used to create more accurate and relevant search results. For example, it could be used to create search results that are tailored to each individual user’s interests and needs. It could also be used to create search results that are more creative and engaging. Generative AI could also be used to create new types of search engines. For example, it could be used to create search engines that are based on natural language processing. This would allow users to search for information in a more natural way, without having to use keywords or phrases. Overall, generative AI has the potential to make search engines more accurate, relevant, and engaging. It could also create new types of search engines that are more user-friendly and natural. However, there are also some potential challenges associated with generative AI. For example, it could be used to create fake news or propaganda. It could also be used to create content that is offensive or harmful. It is important to be aware of these potential challenges, but it is also important to remember that generative AI is a powerful tool that has the potential to make the world a better place. |
ก็บอกตรงๆ เลยว่า ผมชอบคำตอบจาก Google Bard มากกว่าครับ เนื้อหามันตรงประเด็นกว่า มันเริ่มจากว่า Generative AI ทำอะไรได้ จากนั้นก็พูดถึงผลกระทบทั้งในด้านบวกและด้านลบ รอบด้านดีจริงๆ แต่ผมว่ามันไม่ใช่ว่า Bard มันเก่งกว่า ChatGPT หรอกนะ แต่ผมว่าข้อมูลผลการค้นหาที่ได้จาก Google มันดีกว่า Bing ครับ แต่ผมตั้งข้อสังเกตว่า ในขณะที่ Microsoft Bing มันมีการแท็กแหล่งที่มาของข้อมูล ฝั่ง Google Bard นี่ไม่มีการอ้างอิงเลยนะว่าข้อมูลมาจากไหน มีแต่ปุ่มที่เขียนว่า Google it ที่พอกดแล้ว มันจะแสดงให้เห็นว่าคีย์เวิร์ดสำหรับการค้นหาคืออะไร ซึ่งเราสามารถคลิก แล้วไปค้น Google แบบดั้งเดิมได้
ทีนี้ผมลองใหม่ครับ ด้วย Prompt ว่า Suggest me a 4 days 3 nights travel plan in Singapore for me. I will travel from Bangkok to Singapore. I will arrive at Singapore at 14:00 on the first day and will leave Singapore to Thailand at 17:00 on the last day. สิ่งที่ได้มาก็ตามรูปด้านล่างครับ


Microsoft Bing | Google Bard |
---|---|
Here’s a 4 days 3 nights travel plan in Singapore that you can consider: Day 1: 🔹Arrive at Singapore at 14:00 🔹Check-in to your hotel 🔹Visit Gardens by the Bay 🔹Visit Marina Bay Sands 🔹Stroll around Chinatown 🔹Visit the Singapore City Gallery 🔹Take a picture of the Merlion 🔹Watch the Singapore skyline by night Day 2: 🔹Visit Sentosa Island 🔹Visit Universal Studios Singapore 🔹Visit S.E.A. Aquarium Day 3: 🔹Visit Orchard Road 🔹Visit Clarke Quay 🔹Visit Little India Day 4: 🔹Visit Singapore Zoo 🔹Leave Singapore to Thailand at 17:00 I hope this helps! Let me know if you need more information. | Here is a possible 4 days 3 nights travel plan in Singapore for you: Day 1: 🔹Arrive at Changi Airport at 14:00. 🔹Check in to your hotel in the city center. 🔹Take a walk around the Marina Bay area and admire the Singapore skyline. 🔹Have dinner at a restaurant in Chinatown. Day 2: 🔹Visit the Singapore Botanic Gardens. 🔹Take a ride on the Singapore Flyer. 🔹Have lunch at a restaurant in Little India. 🔹Visit the Singapore Art Museum. Day 3: 🔹Visit the Singapore Zoo. 🔹Take a boat ride to Sentosa Island. 🔹Visit the Universal Studios Singapore. Day 4: 🔹Have breakfast at your hotel. 🔹Check out of your hotel and take a taxi to Changi Airport. 🔹Depart from Singapore at 17:00. This itinerary can be customized to fit your interests and budget. For example, if you are interested in history, you could visit the Singapore National Museum or the Asian Civilisations Museum. If you are interested in nature, you could visit the Singapore Botanic Gardens or the Jurong Bird Park. If you are interested in shopping, you could visit the Orchard Road shopping district or the Chinatown Street Market. The total cost of this trip will vary depending on the time of year you travel, the type of accommodation you choose, and the activities you participate in. However, you can expect to spend around $1,000-2,000 per person for a 4 days 3 nights trip to Singapore. Here are some things to keep in mind when planning your trip: 🔹The best time to visit Singapore is during the dry season, which runs from December to April. 🔹Singapore is a very safe city, but it is always a good idea to be aware of your surroundings and take precautions against petty crime. 🔹The official currency of Singapore is the Singapore dollar. You can exchange your currency at banks or currency exchange bureaus. 🔹English is the official language of Singapore, but you will also find many people speaking Malay, Chinese, and Tamil. 🔹Singapore is a very cosmopolitan city, and you will find a wide variety of restaurants and cafes serving food from all over the world. 🔹Singapore is a very popular tourist destination, so it is a good idea to book your accommodation and activities in advance. |
ในแง่ของแผนการท่องเที่ยว ผมว่ามันก็ครอบคลุมสถานที่ท่องเที่ยวหลักๆ ของสิงคโปร์ดี มันต่างกันแค่การจัดตารางว่าวันไหนไปที่ไหนนั่นแหละ แต่เอาจริงๆ นะ ถ้าผมไปจริง ผมไม่ได้ไปวันนึงเยอะที่ขนาดนั้นหรอกครับ แต่สังเกตว่า Google จะคิดเผื่ออะไรหลายๆ อย่างเลย เช่น มีตัวเลือกอื่นไหม ยังไง ใช้งบประมาณเท่าไหร่ และมีอะไรบ้างที่เราควรรู้ ซึ่งเป็นอะไรที่ Microsoft Bing ไม่มี
ผมรู้สึกว่ามันอ่านยาก เลยลองสั่งเพิ่มเข้าไปครับว่า ให้ทำมาเป็นตาราง อันนี้เพื่อทดสอบว่า ในฐานะ Generative AI มันจะสามารถจดจำคำสั่งเดิมของผมได้ไหม จำได้ไหมว่าตอบอะไรผมมา แล้วมันจะสามารถจัดการรูปแบบการนำเสนอของผลการค้นหาให้ผมได้ไหม ผลก็ออกมาตามนี้ครับ


จะเห็นว่าทั้ง Microsoft Bing และ Google Bard ก็สามารถจดจำได้ว่าเคยตอบอะไรผมมา สามารถคุยกับผมต่อได้เลย และแสดงผลเป็นตารางได้หมด แต่มันดูเพี้ยนๆ หน่อย โดยเฉพาะ Microsoft Bing น่าจะเป็นเพราะว่ามันยังไม่สามารถสร้างตารางที่มี HTML ที่ซับซ้อนได้ครับ สังเกตจากที่แท็ก HTML หลุดมาเฉยเลย ส่วน Google Bard นี่ พวก Bullet points ก็หายไป แต่ก็ไม่ได้หมายความว่า Google Bard มันเก่งกว่านะ ก่อนหน้านี้ผมมีให้มันลองสร้างข้อมูลเป็นตารางออกมา มันเละเทะเลยครับ ไม่มีความเป็นตารางเลย
ผมถาม Microsoft Bing ว่าจะเขียนโค้ด Python เพื่ออัปโหลดไฟล์ไปที่ FTP server ยังไง และ Microsoft Bing ก็สามารถตอบมาได้สบายๆ อันนี้ไม่แปลกใจ เพราะ ChatGPT ก็ทำได้อยู่แล้ว ถ้า Microsoft Bing ทำไม่ได้นี่สิแปลก

แต่ในทางกลับกัน Google Bard ตอบไม่ได้ครับ และผลการสนทนาครั้งนึงที่ผมคุยเรื่องการสร้างตาราง Bard ก็ตอบผมกลับมาด้วยว่า I’m not trained for coding yet, so I can’t help with that right now. I can help you write things, brainstorm ideas, or answer other complex questions. Maybe we can try something like that?

ดูๆ แล้ว Google ยังไม่ได้เทรน Bard เรื่องการเขียนโปรแกรมนะ อาจเป็นไปได้ว่า ในแง่ของความสามารถทั่วไป Bard อาจจะยังไม่เก่งเท่า ChatGPT ไหมนะ
วันนี้เอาพอเท่านี้ก่อน เดี๋ยวขอเล่นอีกยาวๆ ก่อนค่อยว่ากันครับ