Amazon Reportedly Training AI With Twice as Many Parameters as GPT-4

Bigger models have a larger capacity for learning world knowledge and the nuances of human language (given that they have access to high-quality training data). For example, GPT-4.5 has a record-high ranking on PersonQA, a benchmark that evaluates hallucinations in AI models. Box CEO Aaron Levie said on X that his company used GPT-4.5 to help extract structured data and metadata from complex enterprise content. Other advantages include larger context windows, which refer to the amount of tokens (pieces of information) the model can process as input and output. For reference, the o1 and o3-mini models in the API have a 200K context length, and GPT-4.5 and GPT-4o have a 128K context length.

The ChatGPT system that exploded in popularity over the last few months was a way to interact with GPT-3.5, and now it’s a way to interact with GPT-4. OpenAI’s new GPT-4 AI model has made its big debut and is already powering everything from a virtual volunteer for the visually impaired to an improved language learning bot in Duolingo. But what sets GPT-4 apart from previous versions like ChatGPT and GPT-3.5?

OpenAI GPT-4 Released: Here’s What’s New And How You Can Try It

Moonshot might have stumbled onto something fundamental about mathematical reasoning that the usual suspects haven’t cracked. The impact of AI on every aspect of life over the coming years and decades will be so profound that the more people are thinking about it in advance, the better. Even with GPT-4 hitting headlines, and obsessing many technophiles, AI still isn’t getting the wider public attention it really deserves. It still gets beaten by populist attacks on transgender people and other somewhat spurious stories, but we must be grateful for whatever progress we can get. GPT-4 integrates steerability more natively than GPT-3.5, and users will be able to change the “classic ChatGPT personality with a fixed verbosity, tone, and style” to something more suited to their needs. “Within bounds,” the team is quick to note, pointing to this as the easiest way to get the model to break character.

Revised timelines and threat estimates

No such description appears in GPT-4’s technical paper, nor do any descriptive phrases give away its architecture or other key features. Despite its relative strengths over GPT-4o and o3-mini, GPT-4.5 isn’t a direct replacement for those models. Compared to OpenAI’s reasoning systems, GPT-4.5 is “a more general-purpose, innately smarter model.” Additionally, it’s not natively multimodal like GPT-4o, meaning it doesn’t work with features like Voice Mode, video or screensharing. Given its deeper world knowledge, GPT-4.5 is also suitable for “LLM-as-a-Judge” tasks, where a strong model evaluates the output of smaller models. For example, a model such as GPT-4o or o3 can generate one or several responses, reason over the solution and pass the final answer to GPT-4.5 for revision and refinement. Given its improved world knowledge, GPT-4.5 can also be a suitable model for creating high-level plans for complex tasks.

However, OpenAI’s GPT-4 Turbo chatbot has knowledge of events up until April 2023. In the wake of Elon Musk’s xAI launching a chatbot boasting access to real time information, this is a key update in the budding Grok vs ChatGPT rivalry. The company has created a lightweight version of Deep Research that is powered by its new o4-mini language model. OpenAI says this variant is “more cost-efficient while preserving high quality.” More importantly, it is available to use for free without any subscription caveat. Several companies have announced deep research features in recent weeks and months that excel in areas such as finance, science, marketing, and academics.

We haven’t heard much about GPT-5 in recent days, but given the speed at which OpenAI has been releasing new models this year, no release date would surprise us. One month after launching its latest AI models, OpenAI is bringing GPT-4.1 to ChatGPT. Stay informed on the top business tech stories with Tech.co’s weekly highlights reel. This is particularly important if you’re using it to generate anything that you’ll be sharing with clients or customers. As you may be aware, ChatGPT and other AI tools like Bard have a tendency to “hallucinate” – so proofreading and fact-checking the content they produce for you is essential, not optional.

• These feelings have been out there for a while, but now we may finally have proof.
• And, the model was so large that OpenAI needed to spread training across multiple data centers to finish in a reasonable time.
• Notably, OpenAI’s announcement doesn’t mention anything about the bias problem, which we found was a recurring problem with ChatGPT in performing creative tasks.
• And our capped-profits structure means we aren’t incentivized to make unlimited returns,” he tweeted.

In order to avoid generating repetitive text, they make some arbitrary tweaks to the probabilities. When a system is tuned to make more tweaks, it is said to have a higher “temperature”. However, it looks like with GPT-4, OpenAI focused more on the safety side of things and emphasis on facts than philosophical ramblings that came from simply summarising the web. The company says the chances of GPT-4 responding to disallowed content is 82% lower, while the probability of answering with facts is 40% compared to models based on GPT-3.5. The level of human input that went into training GPT-4 was also higher, ensuring that the responses sound more natural than the machine-generated repetitive tones that are discernible with ChatGPT. Although ChatGPT was originally described as being GPT-3.5 (and therefore a few iterations beyond GPT-3), it is not itself a version of OpenAI’s large language model, but rather a chat-based interface for whatever model powers it.

Once you’re a paying customer, your access to the new model via ChatGPT will be immediate. The green “Upgrade Plan” button will whisk you to a standard e-commerce sales page, where you’ll need to input the usual information, and cough up $20 per month. Notably, GPT-4 is currently available for developers or paid members through ChatGPT Plus.

It breaks ground in acknowledging the enormous resources marshaled to make the program operate. “Due to our concerns about malicious applications of the technology, we are not releasing the trained model,” said OpenAI. That technical disclosure allowed many researchers to reason about the functioning of the program even if they couldn’t duplicate its construction. ChatGPT has also announced that it will be reducing token prices, “passing on savings to developers” in the process.

“Want to talk about what happened, or do you just need a distraction? I’m here either way,” the chatbot says when powered by GPT-4.5. GPT-4 Turbo is the latest language model to be released by ChatGPT owner OpenAI. It’s more powerful than the previous two language models that were used to power ChatGPT, GPT-4 and GPT-3.5. They are considered the first steps toward the concept of artificial general intelligence (AGI), which some define as a model that can process a query based on novel data that it has not been trained on, and it can produce unique content.

Generally speaking, models with small context windows tend to “forget” the content of even very recent conversations, leading them to veer off topic. GPT-3 was launched in November 2020 and boasted what was then an unheard-of number of 175 billion parameters (analogous to the synapses in a human brain). GPT-4 was released on 14 March 2023, and the number of parameters has not been disclosed. OpenAI has been criticised for reversing its policy of publishing as much as possible about its systems. It replies, not unreasonably, that if these models can cause harm in the wrong hands, it would be silly to make it easier for the bad guys to replicate them.