Skip to main content

Jeopardy! IBM Challenge continues as Watson takes a huge lead

Round two of the Jeopardy! IBM Challenge continued last night as past champions Ken Jennings and Brad Rutter returned to face off against the trivia-spouting supercomputer, Watson. The first round ended with Rutter and Watson tied at $5,000 and Jennings trailing at $2,000. Last night’s showdown played out very differently as the IBM creation pulled far ahead of its meatbag competitors over the course of the episode.

Despite making a few flubs on key questions, a high bid Daily Double and the Final Jeopardy question in particular, Watson ended the second of three rounds with $35,734, more than triple second-place contender Rutter’s $10,400 and miles ahead of Jennings’ $4800. The human competitors are going to have their work cut out for them in the final faceoff tonight, though looking at Watson’s overall performance in round two, this contest is pretty much over.

Recommended Videos

The supercomputer tripped up a couple of times last night, notably when it bid $6435 on and got the response wrong after misreading the clue. “In May 2010 5 paintings worth $125 million by Braque, Matisse & 3 others left Paris’ museum of this art period,” the clue read. Watson replied “Picasso,” naming one of the three other artists whose paintings were stolen (the other two were Amedeo Modigliani and Fernand Leger) when, of course, the response being sought was the rest of the museum’s name, “Modern Art.”

Twice last night the supercomputer repeated a wrong answer delivered by a quicker-to-the-buzzer competitor. Watson also had trouble in general with the art-related clues. Didn’t Star Trek: The Next Generation‘s android Data struggle with such things as well? Computers and creative thinking clearly do not go well together.

Watson also failed to answer correctly during Final Jeopardy, a clue asking them to name which U.S. city has one airport named after a World War II hero and another named after a World War II battle. “Toronto?????” the computer replied, which isn’t even a U.S. city. It bet only $947 though, which kept the final money tally far ahead of the other two competitors. Tonight’s third and final round will give Jennings and Rutter one last opportunity to gain the upper hand, but for now it’s looking like a victory for the machines of the world.

Adam Rosenberg
Former Digital Trends Contributor
Previously, Adam worked in the games press as a freelance writer and critic for a range of outlets, including Digital Trends…
Some older D-Link routers are vulnerable to attack
D-Link Omna 180 Cam HD

A few legacy D-Link routers can be vulnerable to Remote Code Execution (RCE) attacks since the company refuses to send any updates to patch them up, claiming they have reached end-of-life, as recently posted on its announcement page.

The vulnerability is a serious issue since it allows hackers to take control from anywhere in the world and use a stack buffer overflow. This attack sends more data than the buffer size can handle, potentially corrupting critical information like the return address. Thus, hackers can take control of your PC. However, the company did not detail how the threat works, possibly not informing the hackers too much about the issue.

Read more
Nvidia warns gamers of an incoming GPU shortage
The RTX 4090 sitting alongside the Fractal Terra case.

Get ready for yet another GPU shortage, as Nvidia has warned about potential shortages in the current quarter. Although the company’s third-quarter revenue saw a healthy growth, in its recent earnings call, company Chief Financial Officer Colette Kress said that the fourth-quarter revenue is expected to decline sequentially due to supply constraints.

The anticipated GPU shortage could potentially stem from a strategic shift in production to prepare for the next-generation RTX 50-series “Blackwell” GPUs, slated for release in 2025. This supply transition, coupled with surging gaming and professional use-case demand, seems to have left the company in a tight spot. Nvidia acknowledged its struggle to maintain stock for both gamers and enterprise customers, emphasizing its ongoing efforts to expand manufacturing capacity.

Read more
ChatGPT’s latest model may be a regression in performance
chatGPT on a phone on an encyclopedia

According to a new report from Artificial Analysis, OpenAI's flagship large language model for ChatGPT, GPT-4o, has significantly regressed in recent weeks, putting the state-of-the-art model's performance on par with the far smaller, and notably less capable, GPT-4o-mini model.

This analysis comes less than 24 hours after the company announced an upgrade for the GPT-4o model. "The model’s creative writing ability has leveled up–more natural, engaging, and tailored writing to improve relevance & readability," OpenAI wrote on X. "It’s also better at working with uploaded files, providing deeper insights & more thorough responses." Whether those claims continue to hold up is now being cast in doubt.

Read more