Cointime

Download App
iOS & Android

Microsoft or: How I Learned to Stop Worrying and Love ChatGPT

In the last article, I discussed ChatGPT, where I explained what we know and how this is not exactly a threat to Google. Moreover, I explained how some claims have no basis.

OpenAI and Microsoft

The field is actually moving quickly and we can see there is some news. In fact, ChatGPT has been considered an earthquake in tech (ChatGPT sets record for fastest-growing user base). In fact, ChatGPT sparked a rising interest among investors and it is fueling an AI race. As noted by the NT Times:

As ChatGPT has captured the world’s imagination, Mr. Altman has been put in the rare position of trying to downplay a hit product. He is worried that too much hype for ChatGPT could provoke a regulatory backlash or create inflated expectations for future releases, two people familiar with his views said. — source

Anyway, I will start this article by discussing the news relative to Microsoft & OpenAI, which in the last days tightened the relationship.

The first one is that Microsoft is still very interested in OpenAI and investing 10 billion in the company:

Microsoft’s infusion would be part of a complicated deal in which the company would get 75% of OpenAI’s profits until it recoups its investment, the people said. (It’s not clear whether money that OpenAI spends on Microsoft’s cloud-computing arm would count toward evening its account.) — source

If you are interested in the details Fortune is describing them here.

In any case, the partnership between OpenAI and Microsoft is official as published by Microsoft’s blog. Microsoft will provide help to OpenAI with infrastructure (Azure) and implement OpenAI’s model in its products.

As seen before, Microsoft has already collaborated with OpenAI for GitHub copilot. However, Wired raises some skepticism on how there will be the integration of ChatGPT on Microsoft products.

  ChatGPT’s Grandfather (source)  

There are rumors that Microsoft will soon implement ChatGPT in Bing, but as I said in my last article, it is not an easy task. Anyway, as 

Paul DelSignore discusses in this article, these days the future of internet search will be reshaped.

Moreover, OpenAI models will be available in Microsoft Azure as declared by the company. But curb your enthusiasm, as detailed by 

Lorenzo Zarantonello you probably would not access it soon.

Second news, ChatGPT has been improved as declared by OpenAI:

We made more improvements to the ChatGPT model! It should be generally better across a wide range of topics and has improved factuality.

Stop generating: Based on your feedback, we’ve added the ability to stop generating ChatGPT’s response


Apparently, it is more than just news, 

Christianlauer noticed that OpenAI is deeply interested in the idea that its model can be good a coding. OpenAI hired many contractors(over 1000 remote workers in Latin America and Eastern Europe) to improve the coding ability of ChatGPT.

From OpenAI is not the only news, they announced a pro-version:

In this pro-version, which should cost 20 $/month, the model should be always available (even during peak usage), the answers will be faster and you have no message limit. As 

LucianoSphere noted in the last days ChatGPT was often saturated, and OpenAI says that in these cases paid subscribers will have priority.

Moreover, they released a cookbook on GitHub:

screenshot by the author.

And they released an AI classifier for detecting AI-written text (basically, providing the cure after releasing the disease). As noted by 

Henry L. Peck OpenAI is generally adding a watermarkto its product so it can easily be detected what is produced by its models. But would it work with AI-generated text from other models?

Ah, many people are asking when would be released GPT-4? Apparently, there would be a delay.

Google and the rest of the world?

Latest news, Google is testing a potential ChatGPT-like chatbot called Apprentice Bard. A part of the name (which is awesome) we do not much about it. It seems it is based apparently on LaMDA:

“As a result of ChatGPT, the LaMDA team has been asked to prioritize working on a response to ChatGPT,” read one internal memo viewed by CNBC. “In the short term, it takes precedence over other projects,” — source

Rumors report Google’s model will have access to the internet and should cite the sources it uses.

It is not actually the first try, Google was working on a smart chatbot called Meena, but it has been discontinued. Moreover, there was already a chatbot demo based on LaMDA, called Pluto. So, it seems Google is actually working hard to respond to Microsoft & OpenAI.

Google last month has laid off around 12,000 workers but seems that many teams involved in AI research have not been impacted.

Anyway, Google is not the only one thinking about its chatbot. Claude is for example one of these alternatives (here is the research paper). Just a few days ago, Google announced to invest 300 million in Anthropic (which is basically the company behind Claude). Ah, curious detail Anthropic was founded by former OpenAI employees.

Perplexity is another alternative (you can use it for chrome as an extension). Or there are also alternatives specific to a particular domain like BioChatGPT.

Al Anany noticed that even if you are not creating your own ChatGPT you can still use it for your business. In fact, BuzzFeed stocks rose 311% because they announce they will use their ChatGPTfor their news and polls.

As it is known, Jasper is wrapped around GPT3, so as 

Jae Duk Seo notices it is expected that other businesseswill use ChatGPT as an engine.

Schools and academia?

University and schools are worried that students will have now an unfair advantage and school essays are dead:

Now there’s a fresh concern: ChatGPT, an artificial intelligence (AI) powered chatbot that creates surprisingly intelligent-sounding text in response to user prompts, including homework assignments and exam-style questions. -source

The New York City school directly banned it. Australian universities decided that it was safer to go back to pen-and-paper exams to avoid students using AI tools. English lecturers are urging universities to review assessments.

As I discussed in the last article, this fear is probably exaggerated but it is actually time we rediscuss education.

In fact, Giada Pistilli (ethicist at Hugging Face) has shared some thoughts on how education can use these tools. She noted, that once there was a fear of computers, but teachers did not disappear just adapted.

Another interesting point of view is shown by one of the brightest mathematicians of today:

Catalunia decided to not ban ChatGPT but instead educate users about technology (students and professors).

Interestingly, as showed by 

Mircea Iosif ChatGPT has obtained passing grades at prestigious university tests: from the Wharton School of Business in Pennsylvania to law exams at the University of Minnesota.

In the end, as 

JoAnn Ryan suggested here, AI-generated writing is just plain old plagiarism. Thus, in the old times, students would ask someone else to write their essays, today they will use ChatGPT (nothing new).

Interestingly, as 

Curt Mercadante discusses here, there are also concerns if ChatGPT in its responses is plagiarizing sources. Taking into account the fact that the model is parroting what it has seen during the training, just a minimal part of its generated content is similar to an existing source (concerns about breaking copyright law are at the basis of the non-release of Google MusicLM).

and some professors are roasting students who think to use ChatGPT:

Moreover, scientists are worried about the impact of ChatGPT:

Since a chatbot called ChatGPT was released late last year, it has become apparent that this type of artificial intelligence (AI) technology will have huge implications on the way in which researchers work. -source

In a report, scientists discuss that abstract written by the ChatGPT can be not spotted:

“I am very worried, if we’re now in a situation where the experts are not able to determine what’s true or not, we lose the middleman that we desperately need to guide us through complicated topics,” Sandra Wachter, University of Oxford, UK. source

There has been already an outcry when researchers listed ChatGPT as an author (which personally I think is a marketing move). As suggested, researchers should list in the methods they used a language model (as it is done when using any machine learning model).

In an editorial, Nature goes much far claiming that ChatGPT threatens transparent science and we need clear guidelines. Anyway, the fear is shared by Science journal, showing how the whole academia is worried.

Anyway, as noted by 

Rafe Brena, PhD ChatGPT is far from being perfect and if you want to use it for essays and/or science is not a good idea

Other interesting readings

If you are interested in how ChatGPT I suggest two articles: this by 

Opal A Roszell and this very detailed article byMolly Ruby. Moreover, if you are interested in GPT (the language model behind ChatGPT) you can read this article by Storius Magazine.

As 

RW Blackwelsh noted in this story, all platforms are starting to take a stance against ChatGPT. For example, StackOverflow banned it or Medium decided you have to declare if you used an AI tool (Timothy Key thinks ChatGPT will improve Medium, but I am still skeptical)

Stackoverflow post, source Aeisha Aslam noticed that also publications editors have spoken about ChatGPT: for example, The writing cooperative and Towards Data Science banned it entirely. Moreover,Luise Schulz noted that Medium is already full of AI-generated content(articles written by ChatGPT or one line and then the rest of AI-written content). While I agree banning AI-written content is not the solution and how Medium suggests AI-written content should be disclosed, authors should not exaggerate.

There is a discussion about the fact AI replacing jobs in the content industry. For example, 

Irene mmassy suggests here that some works like data entry, content generation, and market service can be automated by AI. I also suggest a nice perspectiveon the topic by Domenic Thomas

Domenic Thomas

I'm seeking the truth between the extreme right and left-leaning ideologies because BOTH sides have lost their minds.

Follow. Still, on the ethics topic, EU regulations are coming and at the moment ChatGPT could not even comply with the GDPR (more on that here by Dennis Hillemann

Among the best suggestions for ChatGPT:

some fails of ChatGPT:

As the field is moving quickly, probably I will have to do soon a follow-up article with the most recent news. What did you think about ChatGPT? Let me know in the comments and share also your best stories on ChatGPT, I am curious to read your ideas.

https://medium.com/data-driven-fiction/microsoft-or-how-i-learned-to-stop-worrying-and-love-chatgpt-2ca1e7b446a3

Comments

All Comments

Recommended for you

  • SBF ordered to forfeit more than $11 billion

    SBF has been ordered to confiscate more than 11 billion US dollars. SBF has now been sentenced to 25 years in prison.

  • Former CEO of FTX and Alameda Research Sentenced to 25 Years in Prison for Fraud and Money Laundering

    Sam Bankman-Fried, the co-founder and former CEO of FTX and Alameda Research, has been sentenced to 25 years in prison for fraud and money laundering. The judge criticized Bankman-Fried's behavior during the trial and deemed a 25-year sentence to be sufficient. Bankman-Fried's sentence may send a message to the crypto industry and there is no possibility of parole, but he may earn "good time" credit for good behavior while incarcerated. Bankman-Fried was found to have misused over $8 billion in customer funds and will be serving time in prison for his actions. The trial emphasized the importance of not using customers' funds without their knowledge or approval.

  • Riding the Bull: Crafting a Winning Call Spread Strategy in BTC

    If you maintain an optimistic perspective regarding BTC, here’s a trading idea customized for the current market environment.

  • SAFE Token Transferability

    If the token transferability proposal reaches a successful vote by April 15, 2024, SAFE token transferability could be enabled on April 23, 2024. View the proposal here.

  • Publisher-Exchanges: Consumer Applications and the Attention Theory of Value

    “The price is the news” is an often used adage in crypto circles any time there is heightened engagement and awareness around a network following fast price action. It has become clear to me that the inverse of this statement is perhaps more profoundly true: “The news is the price.”

  • The Token Is The Product

    There’s an old saying in venture that says “first time founders focus on product, and second time founders focus on distribution”. This describes how product builders often expect that they can achieve growth purely thanks to the quality of their product, rather than by investing energy into creating repeatable patterns that will help them consistently attract attention and users to their product.

  • Ethereum has blobs. Where do we go from here?

    On March 13, the Dencun hard fork activated, enabling one of the long-awaited features of Ethereum: proto-danksharding (aka EIP-4844, aka blobs). Initially, the fork reduced the transaction fees of rollups by a factor of over 100, as blobs were nearly free. In the last day, we finally saw blobs spike up in volume and the fee market activate as the blobscriptions protocol started to use them. Blobs are not free, but they remain much cheaper than calldata.

  • Web3 AI training company FLock raises $6 million in seed funding

    Web3 artificial intelligence training company FLock has raised $6 million in seed funding led by Lightspeed Faction and Tagus Capital. FLock will use these funds to develop its team and build a federated learning-driven artificial intelligence training platform.

  • Prisma: Vault owners need to prohibit delegation of contracts related to LST and LRT

    The LSD stablecoin protocol Prisma Finance stated in a post that for vault owners, please prohibit delegating authorization of the LST contract starting with 0xcC72 and the LRT contract starting with 0xC3eA.

  • MAS: Singapore is working on global first-tier fund tokenization regulation

    Chia Der Jiun, Managing Director of the Monetary Authority of Singapore, introduced some fund tokenization pilots at an event for asset managers. These pilots are part of the Project Guardian and MAS Global Layer 1 (GL1) tokenization plans. Chia Der Jiun emphasized the advantages of tokenization in real-time settlement and process automation, which can improve efficiency and achieve greater customization of funds. UK asset management company Schroders and fund distribution platform Calastone are exploring this as part of the Project Guardian public blockchain trial in Singapore. A recent survey by Calastone showed that 96% of asset management companies in the Asia-Pacific region plan to launch tokenized products within three years. Chia stated that as these Project Guardian pilot projects approach commercialization, MAS is working with the pilot project managers to study the legal and regulatory treatment and impact of tokenized investment funds."